How to appear in Google AI Overviews
Google AI Overviews is the generated summary that sits above the traditional ten blue links on roughly one in five searches, and it has crushed organic click-through rates on the queries it fires for. As of Q2 2025, AIO reaches 2 billion monthly users globally. Inside the queries where it shows, it collapses organic CTR by 61 percent and paid CTR by 68 percent (Seer Interactive's 25-million-impression study). Being cited inside AIO is now the largest swing factor for Google organic traffic. This is the 2026 playbook we use to get Soar clients into AIO citations instead of under them.
What Google AI Overviews are and why scale matters
Google announced AI Overviews at I/O on May 14, 2024 (Google announcement), and the feature reached 2 billion monthly users worldwide by Q2 2025 (TechCrunch). AI Mode added another 100 million in the US and India. No other AI search product has that reach. For most brands, AIO is the first visibility problem to solve, not the fourth.
When AIO appears, the results page works differently. The summary answers the question at the top, the links get pushed down, and users click through less even when they scroll. Any brand not measuring AIO presence on target keywords is measuring organic traffic without accounting for the biggest CTR variable of 2025 and 2026.
When AIO triggers and how CTR actually moves
AIO does not fire on every query. Sistrix 2025 tracking found it on roughly 20 percent of German-language keywords and 18 percent of UK-language keywords (Sistrix). Informational queries get hit hardest. Transactional queries are mostly spared.
Inside those queries, Seer Interactive's September 2025 study of 25.1 million organic and 1.1 million paid impressions (full report) found:
- Organic CTR dropped 61 percent (from 1.76 percent to 0.61 percent) when AIO was present
- Paid CTR dropped 68 percent
- Content cited inside AIO gained 35 percent more organic clicks versus not being cited
- Paid clicks went up 91 percent under the same condition
Ahrefs published a December 2025 analysis showing AIO correlates with 58 percent lower CTR on top-ranking pages (Ahrefs). Pew Research found users click 8 percent of the time with an AI summary versus 15 percent without: a 46.7 percent relative drop.
Every number tells the same story. Ranking number one without being in the AIO costs most of your traffic. Being cited almost recovers it. The optimization target is inside the box, not above it.
How Google actually picks AIO citations
Google's Search AI features documentation is clear on one point: there is no special schema.org markup, no separate meta tag, and no dedicated AIO lever. The guidance for AIO, AI Mode, and traditional Search is the same: follow Helpful Content guidelines, maintain E-E-A-T, publish fresh content, let ranking do its work.
What Google does not publish is the weighting that decides which results get quoted inside an AIO versus which only appear below it. From hundreds of client tests across 2025 and 2026, AIO citations correlate with:
Top-10 ranking on the underlying query. Google pulls from its existing index. If you are not top 10 on the traditional results, you are rarely cited in the Overview.
Passage-level specificity. AIO quotes short passages (one or two sentences) that answer a sub-question directly. Pages with H2 or H3 headings matching the sub-question often get cited for that chunk even when the overall page does not rank first.
Freshness signals. For queries where the answer changes over time, AIO strongly prefers recently updated pages. Five-year-old evergreen content rarely gets cited even when it still ranks.
Author and publisher signals. E-E-A-T is doing real work. Pages with named authors who have topical authority get cited more frequently than posts with no byline or "admin" attribution.
Multi-source corroboration. AIO draws quotes from multiple sources in a single summary. Contradicting the consensus will not get you cited unless you bring strong authority signals.
On schema: Google's docs say no AIO-specific schema is needed, and we have tested that. Article, Organization, Product, and FAQPage schemas still help traditional Search ranking which feeds AIO. The full walkthrough is in schema.org markup for AI citations in 2026.
Gemini 3 and the November 2025 Search deployment
Google released Gemini 3 on November 18, 2025 and deployed it to Search on day one (Fortune). That is notable because previous Gemini releases rolled out on a delayed schedule. Day-one deployment means model improvements immediately show up in AIO quality, citation selection, and summary generation.
AIO behavior changed measurably starting in late November 2025. Baselines taken before that date are not comparable to metrics taken after. Any serious AIO measurement program should note the Gemini 3 deployment as a reset point and re-baseline.
Google-Extended and the user agent question
Google introduced the Google-Extended user agent on September 28, 2023 as an opt-out for brands that wanted to skip training without affecting traditional Search ranking. Blocking it prevents your content from training future Gemini versions. It does not remove you from Search ranking.
Here is the subtlety most brands miss. AIO is generated at inference time from content in Google's existing Search index, not from training data directly. Blocking Google-Extended does not block your content from appearing in AIO. Leave Googlebot allowed and decide on Google-Extended based on whether you want your content feeding future training runs. The full breakdown is in the AI bots robots.txt guide.
What actually moves AIO citation rate
After two years running AIO-focused programs, the interventions that consistently move the metric are familiar. The category-specific twists are where the work gets interesting.
Rank well in traditional Search first. Nothing else matters if you are not top 10. E-E-A-T, Helpful Content, site speed, mobile usability, and internal linking are prerequisites.
Build passage-level answers. Add H2 or H3 sections matching specific sub-questions. Answer each in two or three sentences immediately under the heading. That is the format AIO quotes best.
Refresh on a schedule. For queries where answers change (pricing, features, market data), update at least quarterly and show a visible "last updated" date.
Add named authors with bylines. If your posts are authored by "Admin," change that. Attach a real human with a real credential and link the bio to a page that shows topical authority.
Run prompt-specific content audits. AIO fires on a specific set of sub-prompts for every category. For "best project management tool" there are roughly 15 sub-prompts that trigger AIO answers. Answer each with a dedicated passage or you will never be cited. Methodology overlaps with how to audit your brand's AI search visibility.
How to measure AIO presence and citation rate
Measuring AIO is harder than measuring rank because the feature fires inconsistently and citation data is not exposed through Google Search Console. The working stack in 2026 combines Sistrix, Ahrefs Brand Radar, Semrush AI Toolkit, and Parse. Pick one anchor tool that tracks AIO presence per keyword and one citation tracker that shows whether your brand is quoted inside the Overview.
The key metrics: AIO presence rate, citation rate, passage match (which sentences Google is quoting), and post-AIO CTR on the queries where AIO fires. Post-AIO CTR is the metric most brands do not track, and it is the one that ties AIO visibility to actual traffic. The 90-day measurement cadence we run is documented in the 90-day GEO program.
Conclusion
Google AI Overviews is the most important AI visibility surface for most brands and the one where the work looks most like traditional SEO. Standard SEO matters (E-E-A-T, Helpful Content, speed, freshness, named authors), and AIO layers on top of it (passage-level answers, sub-prompt coverage, quarterly refreshes, corroboration). Brands that treat AIO as separate from SEO duplicate work. Brands that treat it as identical miss the passage layer that decides who gets quoted. The winning playbook runs both together.
How Soar saves you time and money
Most brands are still treating AI Overviews as a classical SEO problem. It is and it is not. Standard SEO matters, but the additional layer is whether your content answers the specific prompts that trigger AIO for your category. We run prompt-specific content audits to identify the gaps between what your content covers and what Google's AIO is actually firing on for your target queries. That audit alone saves most clients 30 to 50 hours of content creation on the wrong topics. It is the difference between publishing six posts that never get cited and publishing two posts that each get cited across dozens of sub-prompts.
The compounding saving is measurement discipline. A DIY AIO program usually tracks rank and assumes clicks will follow. In 2026 that correlation is broken by AIO itself. We run parallel measurement against rank, AIO presence, citation rate, and post-AIO CTR, so clients can see exactly where the interventions are landing and where they are not. The first 90 days of the engagement produce a measurable improvement in citation rate on a defined prompt set, and the reporting cadence continues monthly after that. Request a proposal and we will run your top 20 target queries through Parse, identify which ones already trigger AIO in your market, and propose a 90-day plan to get your brand cited inside them.