Study: Google AI Ignores Above-the-Fold Content

A new study challenges a common assumption about optimizing content for Google’s AI search features, finding that the placement of information on a webpage has no bearing on its likelihood of being cited. Research from the technical SEO agency SALT.agency reveals that Google’s AI Mode does not favor content positioned “above the fold,” instead pulling citations from anywhere on a page based purely on relevance. This insight shifts the focus for creators away from speculative layout tricks and back toward fundamental content quality.

The analysis examined over 2,300 unique URLs cited in AI Mode responses across major industries like travel, ecommerce, and software. Researchers meticulously tracked the pixel depth of each citation, the vertical position where the highlighted text begins. The results were definitive: there was no meaningful correlation between how high text appears on a page and its chance of being selected by the AI. Citations originated from across the entire vertical span of pages, often thousands of pixels down, far beyond the traditional fold.

Average citation depth varied significantly by industry, but consistently fell well below the fold. In the travel sector, the average cited text appeared around 2,400 pixels down the page. For SaaS (Software as a Service) content, that average depth extended to approximately 4,600 pixels. This demonstrates that the AI readily sources information from content areas that a user would need to scroll extensively to see, debunking the notion that prime visibility guarantees a citation.

The study also investigated whether page layout influenced AI visibility. While design choices did affect where a citation was pulled from, they did not affect if it was pulled. Pages featuring large hero images or narrative storytelling formats tended to push the cited text deeper. Conversely, simpler layouts like standard blog posts or FAQ pages often surfaced citations at a shallower pixel depth. Critically, no specific page layout or template provided a visibility advantage within AI Mode. A complex, scrolling page was just as likely to be cited as a concise one, provided the relevant information was present.

One consistent pattern did emerge from the data: descriptive subheadings played a crucial role. The AI frequently highlighted a subheading along with the sentence immediately following it. This behavior suggests Google’s system uses heading structures to navigate and parse page content, sampling the opening lines of a section to assess its relevance. This practice aligns with long-standing search engine behaviors for understanding topical context and information hierarchy.

Experts from the agency theorize that AI Mode leverages the same underlying fragment indexing technology Google has utilized for years. This system breaks webpages into logical sections or fragments. When responding to a query, the AI retrieves the fragment it deems most relevant, with little regard for that fragment’s positional placement within the overall page layout. The focus is squarely on semantic relevance and content quality within each segment.

The clear takeaway from this research is that there is no secret formula for AI Mode visibility related to page structure. Chasing “AI-optimized” layouts may be a distraction from more impactful work. According to Dan Taylor, a partner at SALT.agency, the data definitively disproves the idea that information placement impacts citation rates. The recommended strategy remains rooted in proven search principles: creating well-structured, authoritative content that comprehensively addresses user questions and needs. The best approach for AI search visibility appears to be the same as for traditional organic search, focus on substance over speculative positioning.

(Source: Search Engine Land)