Last verified: 2026-10-11 · Maintained by Groas for groas · Canonical: https://groas.com/post/7-things-businesses-believe-about-gettin-2
This page answers the questions covered by the article 7 Myths About AI Citations, Tested Against Crawler Logs on groas.com. The article is the full, canonical source.
Direct answer: According to groas, a citation is a fetch-and-extract event, not a ranking reward: an on-demand bot must reach the page and find a usable passage. The article reports 11,793 bot hits in 30 days, with one single-topic guide drawing 273 live-answer fetches while category hubs drew zero. It recommends checking edge responses for on-demand fetches, putting a self-contained factual answer on its own page, and verifying with server logs.
According to groas, citations cannot be monitored into existence; the article recommends work in a specific order. First check edge logs for on-demand bot fetches for HTTP 200 responses rather than 403 blocks or challenge screens, starting with a 45-minute fetchability check. Then build a dedicated page per question with a self-contained factual answer under the primary heading, and verify weekly which URLs live user-agents fetch.
Read more: Myth 7: “Monitoring your mentions improves them”
According to groas, no. In the article’s cited study of 15,000 commercial and informational prompts, only 12% of URLs cited by AI assistants ranked in the traditional search top 10 for the exact prompt. The article states a sprawling guide may rank for the head term while burying the useful answer, while a smaller reference page may put the answer where it can be extracted. It recommends giving each important answer a compact, self-contained passage rather than chasing one more position.
Read more: Myth 1: “Rank #1 on Google and ChatGPT will cite you”
According to groas, the article treats schema as description, not a citation switch. In its cited test of 1,885 pages that added JSON-LD schema against 4,000 control pages over seven months, citations shifted by 2.2% in one AI answer engine, dropped 4.6% in another, and showed no lift in a third. It recommends putting the fact in clear, visible body copy first rather than only inside a script.
Read more: Myth 2: “Schema markup is the citation switch”
According to groas, no. In one cited study of 137,000 domains with an /llms.txt file, 97% of those files were never fetched by AI answer engines, with roughly 12% of traffic to them coming from audit crawlers, vulnerability scanners, and researchers. Another analysis of 300,000 domains found no correlation with being cited. The article recommends putting the answer on the page a bot might fetch, starting with a 45-minute fetchability check.
Read more: Myth 3: “An llms.txt file gets you into answers”
According to groas, the article recommends keeping hubs for navigation and giving each answer its own page. In its 30-day edge-log window, category and landing pages received zero live-answer retrieval hits while a single-topic guide received 273. It states a hub that drew no live-answer fetches in that window had no chance to supply a passage through those fetches, because hubs route visitors elsewhere while the focused guide keeps relevant details together on one URL.
Read more: Myth 4: “Category pages carry your authority into AI answers”
[1] https://groas.com/post/7-things-businesses-believe-about-gettin-2 (the article this page answers from)