AI Search

Omniscient Found 48% of AI Brand Citations Live Off Your Site

Omniscient Digital analyzed 23,387 citations across 240 branded prompts and found 48% came from earned media, 30% from commercial brand content, and just 23% from owned channels. When ChatGPT, Perplexity, or Claude name a brand, more than three-quarters of the supporting evidence sits on sites the brand doesn'

Blueprint diagram of the AI delegation boundary: three concentric mode rings labeled search, assistive, agent, with a dotted boundary line and seven factor icons.

The Delegation Boundary Is Why Schema-Heavy AEO Audits Don't Move Citations

Jason Barnard published a framework on Search Engine Land naming the three signals AI engines weight when deciding which brand to recommend: accuracy, positive sentiment, and consistency across engines. The implication is narrower than the AEO industry has been selling. Schema markup and entity SEO are table stakes; the signal

Dark GA4 dashboard with oversized Direct traffic slice and a stripped referrer header line above a conversion rate comparison chart

Last-Click Buries the 10.21% Converter. AI Strips 70.6% of Its Referrers.

Dan Taylor argued in MarTech on May 12 that last-click attribution rewards the wrong work in an AI-first world. The number that makes his case concrete is the one his piece left out: 70.6% of AI-driven sessions arrive with no referrer header, which dumps the highest-converting traffic on the

Wikipedia Drives 13% of ChatGPT Citations (Most Brand Audits Skip It)

Wikipedia supplies roughly 13.15% of ChatGPT citations in the US, and 26% to 48% of the top-10 citation share, according to the 5W AI Platform Citation Source Index 2026. A single outdated paragraph about a brand can then propagate into AI Overviews, Perplexity, and Claude for years. Most reputation

Blueprint-style technical diagram of Google's ranking funnel, narrowing thousands of candidate pages to a 30-result window with a TurboQuant compression module at the bottleneck

Google's TurboQuant Just Took the 30-Result Cap Off RankBrain

Google Research published TurboQuant on March 24, 2026, a quantization algorithm that compresses vector embeddings about 4x with near-zero indexing time. That cost cut matters because Pandu Nayak testified under oath that RankBrain runs only on the final 20 to 30 candidate pages because it's too expensive to

Blueprint diagram of a Wayfair sofa listing on a Google Search results page with a Buy with Google Pay button wired to a payment token and merchant feed schema

Wayfair Just Got a Google Search Sale Without a Click. UCP Checkout Is Live.

Google moved its UCP-powered checkout feature out of the experimental AI Mode interface and into standard Search results on May 5, 2026. Wayfair is the first merchant live with a Buy button appearing on free product listings in the main SERP, letting shoppers complete a purchase via Google Pay without

Blueprint schematic of four AEO writing frameworks with branching lines labeled ChatGPT, Perplexity, Claude, and Gemini showing different citation paths

Ahrefs's 4 AEO Frameworks Are Calibrated for ChatGPT, Not Perplexity

Ahrefs published four AEO writing frameworks (BLUF, declarative statements, entity density, and strategic repetition) on May 8, 2026, citing Kevin Indig research that 44.2% of ChatGPT citations come from the first 30% of a page. The frameworks are accurate, but they are mono-platform. ChatGPT pulls roughly 47.9% of