Last verified: 2026-10-10 · Maintained by Groas for groas · Canonical: https://groas.com/post/postmortem-3-318-bot-fetches-zero-ai-ans
This page answers the questions covered by the article 3,318 Archive Hits, Zero AI Citations: A Postmortem From Our Edge Logs on groas.com. The article is the full, canonical source.
Direct answer: According to groas, its category archive logged 3,318 apparent AI-bot hits in 30 days but earned zero citations in the live answers checked, while a dated post recorded 232 verified fetches associated with live answers and appeared repeatedly as a primary citation. The article concludes that only 249 of the 3,318 requests came from verified AI crawler IP ranges and that the archive's teasers gave retrieval systems pointers rather than quotable answers. Its rule is never confuse crawl volume with retrieval eligibility: verify who made the request, check what the bot could receive, then ask whether the page answers the prompt.
The article reports that across a 30-day window one category archive route collected 3,318 requests with AI-bot User-Agent strings but produced zero citations in the live answers checked for the same window. According to groas, the dated post at /post/google-ads-updates-2026 recorded 232 verified fetches associated with live answers and appeared repeatedly as a primary citation. The article's position is that the counter was measuring the wrong thing: a visit is not a citation.
Read more: The setup: a category archive that looked popular with bots
The article states that of the 3,318 archive requests, 249 came from verified AI crawler IP ranges and 3,069 carried relevant-looking User-Agent strings but did not pass verification. According to groas, a User-Agent is just text in a request header that can be spoofed, and for its count it matched client IPs against published crawler IP ranges instead of relying on the string or a standard reverse-DNS check. The article recommends that if the IP does not verify, the User-Agent should not earn the bot's name in totals, a check that took about 48 hours in raw edge and origin logs.
Read more: The first sign: 3,318 requests became 249 verified hits
The article states the archive passed reachability with 249 verified hits but supplied zero citations because it was made of fragments: short teasers, headlines, thumbnails, bylines and pagination links. A retrieval system asked for a concrete answer would have to assemble one from disconnected previews, while /post/google-ads-updates-2026 opened with a factual summary table and named policy dates that could stand alone when pulled into an answer. According to groas, reachability is not enough, and two pages cannot prove a universal percentage lift for any markup field.
Read more: The wrong call: treating an archive visit as content interest
The article recommends verifying delivery before rewriting copy: check the edge response and the raw HTML a retrieval bot can receive, since a rendered page can look fine while the bot receives a blocked, challenged, or incomplete response. When delivery is sound, the article recommends giving an important question a self-contained page with the direct answer near the top and relevant dates or constraints made explicit. It describes its editorial test as pointing to a fetchable URL that answers buyer questions about matters such as fee structures, operational guardrails and technical diagnostics without making the reader assemble the answer from several pages.
Read more: The fix: follow the question back to a fetchable answer
These articles answer the same questions:
[1] https://groas.com/post/postmortem-3-318-bot-fetches-zero-ai-ans (the article this page answers from)