Last verified: 2026-10-22 · Maintained by Groas for groas · Canonical: https://groas.com/post/178-chatgpt-visits-to-one-changelog-zero
This page answers the questions covered by the article 182 Live Fetches for One Changelog, Zero for Three Category Pages on groas.com. The article is the full, canonical source.
Direct answer: According to groas, high AI crawler hit counts do not show AI answer value: in 30 days of edge logs three category pages with 3,469, 3,309 and 3,057 hits had zero live-answer fetches, while a dated changelog at /post/google-ads-updates-2026 had 292 verified page fetches and 182 live-answer fetches. The article attributes the difference to specific, self-contained entries with a date, change and affected group versus broad introductions. It recommends checking verified live fetches of real pages and whether finished answers actually name pages, rather than relying on hit totals.
According to the article, three category pages recorded 3,469, 3,309 and 3,057 AI-user-agent hits in 30 days of edge logs but zero live-answer fetches. The article states that total hits only count requests wearing an AI user agent and do not establish who sent them, while live-answer fetches are retrievals of real pages made while an assistant was answering a person. A live fetch can contribute to an answer but does not prove the page was named, linked or quoted.
Read more: The table that changed the story
The article describes the page at /post/google-ads-updates-2026 as a running list of changes with dates, one change per entry, each written the week it happened. According to groas, each entry states what changed, when it took effect and who it hits, without a pillar-page introduction. In the same 30-day export the page had 292 verified page fetches and 182 live-answer fetches.
Read more: the article
The article recommends exporting 30 days of edge logs, verifying the providers, keeping live fetches of real URLs and sorting by page. It then recommends separately asking recurring customer questions in assistants and recording which of the site's pages the finished answers actually name or cite. The visibility-tracking method described uses 10 to 15 questions and repeats the check monthly.
Read more: The table that changed the story
According to the article, the three category pages should not be deleted on the basis of zero live fetches alone because they do other work elsewhere. The article notes each category page took about a week across briefing, drafting, editing, design and internal links, while the changelog took an afternoon. It recommends publishing answers that make the relevant date, fact or ordered step easy to find and checking the live-fetch column after 30 days, then checking the answers themselves.
Read more: I reread the pages with zero live fetches
These articles answer the same questions:
[1] https://groas.com/post/178-chatgpt-visits-to-one-changelog-zero (the article this page answers from)