Kevin Indig's halftime report shows Google now owns 30% of AI search (Claude 10%), while 68% of all Google searches end zero-click — the open-web referral model is broken.
Why it matters: The traffic loss is now quantified — marketing operators need an AI citation strategy, not an expectation of blue-link volume returning.
Analysis of 16M brand appearances found Perplexity worst (52% unnamed), Copilot best (19%); citations and brand mentions must be tracked as separate KPIs.
Why it matters: A high citation count can mask near-zero brand awareness in AI answers — visibility measurement needs to separate 'cited' from 'named.'
geoSurge's study of ~4,000 AI responses found known brands in 55.7% of queries vs. 17.4% for unknowns; 63% of brand-specific searches routed to one of the model's top-5 familiar companies.
Why it matters: Brand-building is now an AI-visibility investment — content quality alone won't compensate for a familiarity gap baked into model training data.
Feed-sourced data beat page-scraped data 100%-to-0% on completeness; 83% of ChatGPT product carousels matched Google Shopping feed entries; AI-referred traffic converted 42% better than non-AI traffic by March 2026.
Why it matters: Feed hygiene — GTINs, accurate pricing, conversational attributes — is now table stakes for AI commerce visibility alongside product page SEO.
One agency found 6.7M annual Google Search impressions from their YouTube channel that never appeared in web-only GSC reporting; videos ranked for queries the website couldn't reach.
Why it matters: Organic reporting is incomplete without multi-platform attribution; content planning should map video and social topics to the same search demand the website targets.