AI Search Visibility
AI Overviews, ChatGPT search, GEO, and LLM optimization.
-
Google AI Mode adds five link types that complicate attribution
Google added five new link types to AI Overviews including inline links and discussion previews. Attribution now works differently than traditional search results.
-
Google tests Gemini 3 model selector in the search bar
Google is testing Pro, Fast, and Auto model selectors for Gemini 3 in the search bar, which could fragment AI Overview citation tracking across model variants.
-
Schema markup does not influence LLM parsing
Schema markup does not help LLMs parse content because transformer models read token sequences, not markup. Vendors are misrepresenting how AI engines actually work.
-
URL paths are semantic inputs for RAG pipelines, not just SEO
URL path segments now function as semantic inputs for RAG systems and LLMs, not just SEO signals. Descriptive paths improve AI retrieval and citation accuracy.
-
Google's query fan-out splits AI queries against classic search
Google breaks AI queries into sub-queries matched against classic search results. Your pages need to rank for decomposed fragments, not complex full questions.
-
ChatGPT free tier triggers web search in only 10.8% of queries
Free ChatGPT uses live web search in only 10.8% of queries versus 47.4% for paid, leaving users vulnerable to stale data and misinformation.
-
LLMs misrepresent brands at training, retrieval, and generation
LLMs misrepresent brands across training, retrieval, and generation stages. Strengthen Schema.org markup and audit third-party references to improve entity accuracy.
-
Bing Copilot test shrinks citation links to superscripts
Bing is testing smaller superscript citations in Copilot Search instead of full-line links, potentially reducing click-through rates from AI answers to source sites.
-
Semrush playbook targets SaaS citation failures in AI search
SaaS sites lose AI citations when product pages use client-side schema and inconsistent naming, and Semrush published an eight-step playbook to fix extraction.
-
Top-ranking sites still get skipped in AI search citations
Pages ranking first in traditional Google results often get skipped in AI Overview and Perplexity citations, and no unified metric tracks AI visibility yet.
-
Fabricated Google core update ranked in Search and AI Overviews
A fabricated Google core update ranked on page one and appeared in AI Overviews. The test shows misinformation spreads through search without fact-checking layers.
-
AI search scores passages, not pages, killing pillar content
AI search engines score individual passages with cross-encoder models, not whole pages, making traditional pillar content strategies less effective for citations.
-
Botify launches AI Visibility, replaces rank with mentions
Botify's AI Visibility tool tracks mention frequency across ChatGPT, Perplexity, and AI Mode, replacing position-based metrics that don't apply to AI search.
-
AI Overview citations now diverge sharply from top 10 rankings
AI Overview citations now diverge sharply from top 10 rankings, with only 38% overlap versus 76% seven months ago. Query fan-out mechanism explains the shift.
-
Pages ranking in Google can be invisible to AI search
Pages ranking in Google can be invisible to AI search due to differences in JavaScript rendering and crawl access. IPullRank's audit framework identifies the gap.
