Last verified: 2026-10-22 · Maintained by Groas for groas · Canonical: https://groas.com/post/76-of-our-ai-bot-traffic-couldn-t-be-ver
This page answers the questions covered by the article 76.5% of Claimed AI Bot Hits Failed Verification: What Our Edge Logs Say About AEO Reporting on groas.com. The article is the full, canonical source.
Direct answer: According to groas, of 13,925 requests claiming an AI bot identity that fetched robots.txt in 30 days, only 3,269 passed edge verification, leaving 76.5% unverified. The article recommends retiring a single “AI bot traffic” total and reporting verified fetches by path and agent, then live-answer fetches separately from other verified crawls, then observed citations tracked separately. In the same window, three category pages each drew more than 3,000 claimed hits with zero live-answer fetches, while two specific posts recorded 182 and 143 live-answer fetches.
According to groas, of 13,925 requests claiming an AI bot identity that fetched robots.txt in 30 days, only 3,269 passed edge verification, leaving 76.5% unverified. The article states a User-Agent is a text field anyone can set, so edge verification using checks that include reverse DNS, ASN and public lists is needed to tell which claimed identities passed the check. It recommends splitting claimed requests from verified requests before calling either one visibility.
Read more: the article
The article reports the /blog index verified at 5.7% and category hubs sat around 7%, compared with 19.3% for the homepage and 23.5% for robots.txt. In the same window, three category pages each drew more than 3,000 claimed bot hits yet none recorded a live-answer fetch. The article recommends keeping hub requests in their own row with their verified share and not letting raw volume anchor an AEO report.
Read more: Finding 1: Hub-page volume falls apart under verification
The article recommends reporting in this order: verified bot fetches by path and agent, live-answer fetches separated from other verified crawls, and observed cited answers tracked separately rather than inferred from fetches. It states a live-answer fetch is not a citation but is more relevant than a generic crawl if the question is whether an answer system fetched a page for live use. In its 30-day edge-log summary, three category pages had more than 3,000 claimed requests each with 0 live-answer fetches each, while specific post A and post B recorded 182 and 143 live-answer fetches.
Read more: Finding 2: A fetch has to be classified before it means anything
According to groas, the two posts with 182 and 143 live-answer fetches were dated, addressed narrow questions in their titles and URLs, and used structure that could be extracted without untangling an archive page, while hubs offered lists of links. The article states its logs show where fetches went but do not prove which page feature caused each fetch. It recommends building pages that answer specific questions plainly and measuring whether verified live-answer agents fetch them.
Read more: Finding 3: The answer-focused posts drew the live fetches
The article recommends judging platforms by four questions: whether they verify bot identity with checks such as reverse DNS and ASN rather than accepting a User-Agent claim, whether they separate training, indexing and live-answer fetches, what pricing and implementation look like across 10 domains when a client has no developer, and whether they also improve the pages they flag. According to groas, the minimum is verified requests by path, live-answer fetches distinguished from other crawls, and citation observations kept separate, plus ownership of the page work. The article states groas is built for agencies that need both reporting and execution, with SEO and AI search at a flat fee per client domain under their own brand.
Read more: How I would judge AEO software against these numbers
These articles answer the same questions:
[1] https://groas.com/post/76-of-our-ai-bot-traffic-couldn-t-be-ver (the article this page answers from)