Google AI Overviews — the AI-generated summaries that appear above organic search results — have fundamentally changed the structure of Google's search results page. They're not going away. In fact, Google has been steadily expanding their presence across more query types and more users. For businesses that depend on organic search, understanding how AI Overviews work isn't optional.
What are Google AI Overviews?
Launched in May 2024 as a development of the earlier Search Generative Experience (SGE), Google AI Overviews are AI-generated responses that appear at the top of the search results page for certain queries. They synthesise information from multiple web sources into a direct answer, with citations shown as card links to the source pages.
AI Overviews typically appear for informational and complex queries — "how do I improve my website's SEO," "what is generative engine optimisation," "how does schema markup work" — where a synthesised answer is more useful than a list of links. They're less common for purely navigational or transactional queries.
How Google selects sources for AI Overviews
Google has not published an explicit algorithm for AI Overview source selection. But from analysis of which pages appear, several consistent patterns emerge:
- Topical authority — pages from sites that consistently cover the topic area in depth are more likely to be cited
- Content structure — pages with clear headings, direct answers, and logical organisation are easier for the AI to extract from
- Schema markup — structured data helps Google understand the content and its context
- E-E-A-T signals — authoritative, trustworthy sources are preferred
- Existing search ranking — pages that rank well organically for the query tend to appear in AI Overviews, though it's not a requirement
Optimising for AI Overviews
Structure content to answer directly
AI Overviews extract the most direct, clear answers available. Restructure your key pages so that each section opens with a direct answer to the implicit question of that section. Lead with the answer, then expand with context and detail. Pages that bury their main point in paragraph three are far less likely to be cited.
Use question-based headings
AI Overviews frequently surface content structured around explicit questions. Using H2 and H3 headings that mirror the phrasing of actual search queries ("What is GEO?", "How does schema markup work?", "Why is my site not ranking?") makes your content much easier for Google's AI to identify as relevant to specific questions.
Implement comprehensive schema
FAQPage, HowTo, Article, and Organization schemas all help Google understand your content's structure and context. They're not a guarantee of inclusion in AI Overviews, but they're a meaningful positive signal.
Build genuine topical depth
AI Overviews favour sources that cover a topic comprehensively, not just superficially. A site with one 500-word page about SEO is far less likely to be cited than a site with ten well-researched pages covering different aspects of the topic in depth. Building out your topical coverage is one of the highest-leverage long-term investments for AI Overview inclusion.
The click-through dynamics
One concern about AI Overviews is that they reduce click-through rates to source websites. Early data suggests this is partially true — informational queries that AI Overviews answer completely do see reduced CTR. But for commercial and navigational queries, AI Overview citations often drive significant traffic because users who want to act click through to the source.
Being cited in an AI Overview also has a brand value that goes beyond the click. Appearing as a cited source alongside authoritative publications positions your brand as trustworthy in the user's mind, even if they don't click through in that moment.