Bing Rewrites SEO Rules as AIO Citations Diverge
▼ Summary
– The overlap between pages ranking in Google’s top 10 organic results and those cited in AI Overviews has dropped sharply, from 76% to about 38% in seven months.
– AI Overviews now appear on nearly half (48%) of all tracked queries, with especially high frequency in verticals like education, B2B technology, and healthcare.
– Google removed its “Design for accessibility” JavaScript SEO guidance, stating the advice to test with JavaScript disabled is now outdated.
– Microsoft’s rewritten Bing Webmaster Guidelines explicitly detail how meta directives like NOARCHIVE control content appearance in Copilot’s AI answers and define new abuse categories like prompt injection.
– The tools and data for measuring AI search visibility are improving, but the strategies for traditional organic rankings and AI visibility are becoming increasingly distinct.
The landscape of search engine optimization is undergoing a significant transformation, driven by the rapid integration of artificial intelligence into core search experiences. New data reveals a widening gap between traditional organic rankings and visibility within AI-generated answers, forcing marketers and webmasters to adapt their strategies. This week’s developments highlight critical shifts in how sources are selected for AI features, the expanding reach of these tools, and updated guidance from search engines themselves.
A fresh analysis from Ahrefs indicates that ranking in Google’s top ten results is a far weaker predictor for earning a citation in an AI Overview than it was just seven months ago. Their study of millions of data points found that previously, 76% of cited pages also held a top-ten ranking for the same query. That figure has plummeted to 38%. Citations are now almost evenly split between pages ranking 11 through 100 and those beyond the 100th position. This suggests Google’s systems are casting a much wider net when sourcing information for its AI answers, potentially due to an upgraded query “fan-out” process that breaks a single search into multiple sub-queries. YouTube has solidified its position as the most frequently cited domain overall in AI Overviews, experiencing substantial growth over the past half-year.
Concurrently, the presence of these AI features has exploded. Independent research shows AI Overviews now trigger on nearly half of all tracked searches, marking a 58% year-over-year increase. Their prevalence is particularly dominant in specific sectors; for education and B2B technology queries, they now appear over 80% of the time. However, classic “ten blue links” results still define the entire search experience for 52% of queries, underscoring that organic rankings remain vital. When AI Overviews do appear, they consume considerable screen space, often pushing the first organic result below the fold on a standard desktop monitor.
In a notable documentation change, Google has removed its longstanding advice for developers to test websites with JavaScript disabled for accessibility and SEO. The company stated this guidance is now “out of date,” clarifying that modern assistive technologies and Google’s own crawler effectively handle JavaScript content. While the page still notes an additional rendering step for such content, the removal of this specific warning reflects the evolution of web technology and search engine crawling capabilities. It’s important to remember this update applies specifically to Googlebot, as other crawlers used by different AI platforms may still process JavaScript differently.
Perhaps the most direct response to this new era comes from Microsoft. Bing has comprehensively rewritten its Webmaster Guidelines to explicitly address AI experiences like Copilot responses and grounding results. The new guidelines detail how specific meta tags influence AI visibility. For instance, they clarify that the NOARCHIVE directive not only prevents caching but also stops content from being used in Copilot’s AI answers. The rewrite also introduces new abuse categories, renaming “Keyword Stuffing” to “Keyword Stuffing and Artificially Engineered Language” and adding a dedicated section for “Prompt Injection and AI Manipulation.” This provides webmasters with documented controls for managing their presence in AI-driven search results, a level of specificity not yet matched by Google’s public documentation.
The collective message from these updates is clear. The strategies for achieving traditional high rankings and for securing visibility within AI summaries are diverging. The tools and data for measuring AI search performance are becoming more accessible, as seen with Bing’s new AI Performance dashboard, yet the relationship between classic SEO and AI citation is becoming less predictable. Success now requires understanding a dual landscape: competing for clicks in organic listings and competing to be the authoritative source summarized and cited by AI.
(Source: Search Engine Journal)
