Aleyda Solis's analysis of the May 2026 Core Update reveals that Google prioritized source type fit and user intent over raw authority, causing significant visibility swings across reference brands, forums, ecommerce, and marketplaces.

Search Engine Land reverse-engineered Google's Geostore system to expose 72 Oyster Rank signals, 793 data sources, and a multi-stage ranking pipeline that goes far beyond proximity.

Most B2B sites let AI agents extract content, but pricing pages consistently fail—driving agents to third-party sources instead—because of opacity, poor machine-readability, or access friction.

Google's new TurboQuant compression technique could eliminate the memory and processing barriers that currently limit vector search to only the top 20–30 results, potentially enabling massive-scale semantic ranking and more personalized AI Overviews.

AI search engines struggle with international SEO due to semantic collapse, geo-drift, and hreflang limitations—Gianluca Fiorelli outlines the problems and solutions.

Search Engine Journal's audit of 50 major websites identifies technical signals that matter for AI search visibility but are being overlooked by most SEOs.

Google's Job Indexing API only confirms notification receipt—not crawling, indexing, or rankings—and approval timelines have become unreliable.

Google Search Console's new Platform properties feature lets SEO teams see exactly how YouTube, Instagram, and TikTok content ranks in Google Search—often unexpectedly well.

Google recommends blocking filter URLs via noindex or robots.txt, but most major e-commerce sites use canonicals instead—and few test which approach actually performs better.

A Chrome extension experiment demonstrates that local AI models like Gemini Nano can handle certain SEO workflows without requiring cloud-based frontier models.