Most published AI-search statistics are borrowed, US-centric, and already stale. cloro monitors ChatGPT, Perplexity, Gemini, Copilot, Google AI Overview, AI Mode, and Google Search continuously, so these studies come from first-party measurement rather than estimates. Each one states its sample, method, and date, and is free to cite, so you can size an opportunity or set a benchmark without running the measurement yourself.
Machine-issued queries reached 18.9% of one B2B domain's Google query impressions in July, up from 3.1% in March — 140,616 impressions since February with zero clicks in every single month.
Across a balanced 735-prompt basket, four of five engines cite on 98–100% of answers; Gemini cited on just 33% in July 2026, and has since climbed. Citation Presence Index: 86/100.
Measured trigger rates run from 67% (Germany) to 93% (US), though each market runs its own prompts; ChatGPT backs its answers with ~12 sources to Gemini's ~1.
Domains that block GPTBot are cited far less by ChatGPT — median citations-per-appearance 0.003 vs 0.417 — and the effect is engine-specific.
Commercial and voice queries trigger an AI Overview ~98% of the time; local restaurant queries just ~3%.
On a fixed US query set, AI Overview presence fell from 96.7% to 89.4% between April and August 2026.
The #1 organic result changes on 31.7% of consecutive days.
Gemini cites a source on just 41% of answers vs 97%+ for ChatGPT and Copilot; YouTube, Google, and Reddit are the most-cited domains in AI search; and Google AI Mode draws 58% of its citations from only 10 sources.
Which brands ChatGPT, Perplexity, Gemini, Copilot and Google AI Mode recommend across twelve software categories. Re-measured every Monday, with every weekly snapshot kept for citation, so you can see a brand gaining or losing ground.
These studies run on the same API that monitors your brand across ChatGPT, Perplexity, Gemini, Copilot, and Google AI Overview. Run any of these cuts on your own domain and your competitors'. Start with 500 free credits a month.
