getting cited and getting recommended are two different things, and only one of them sells a coach's service.

Google's help documentation for the Generative AI performance report is explicit that it excludes "clicks, CTR, position, or query data". It reports impressions in AI Overviews and AI Mode, broken down by page, country, device and date. It is still rolling out to a subset of properties, UK first, and will not appear if a site has too few AI impressions. The same release added a toggle that removes a site from AI Overviews, AI Mode and Discover AI without affecting normal Search ranking.

Cloudflare now splits bots into Search, Agent and Training. From 15 September, Training and Agent bots are blocked by default on ad-monetized pages. Search bots stay allowed everywhere. The new defaults apply to new domains, new sites of existing customers, and all existing free-tier customers. Cloudflare also notes that multi-purpose crawlers including Googlebot, Applebot and Bingbot get blocked for anyone who selects "block Training".

1,094 US categories tracked monthly in ChatGPT from January to June 2026, five prompt variants each, 283,215 domain-category citation observations and 76,493 brand-mention observations. In categories close to a brand's core, brands were cited as a source in 74% of appearances, mentioned by name in 44%, and both in 34%. In distant categories that fell to 50% cited, 25% mentioned, 9% both. Of citations in the least related categories, only 18% also carried a named brand, against 46% in the most relevant. On repetition: a brand appearing in only one of a category's five prompts saw its mention share drop, and the penalty did not clear until it appeared in at least three of five.

137,210 domains with traffic in May 2026, of which 38,360 published a valid llms.txt. 97% of those files got zero requests in the month. Of the 3% fetched, 96% of requests were bots, AI bots of all kinds were 19.5% of that sliver, and AI retrieval bots specifically (OAI-SearchBot, PerplexityBot, Claude's search crawler) made only a couple of hundred fetches between them. Zero requests came from AI bots looking for llms.txt files that did not already exist. Caveat Ahrefs states: their customer base is more SEO-aware than the web, so 28% adoption is high.

1,885 pages that added JSON-LD between August 2025 and March 2026, matched difference-in- differences against control pages with similar citation histories, measured 30 days either side. AI Overviews −4.6%, AI Mode +2.4%, ChatGPT +2.2%, the last two indistinguishable from noise. The limitation matters: every tested page already had 100+ AI Overview citations, so this says schema does not lift pages that are already being cited. It does not prove schema is useless for a business starting from zero.