Last verified: 2026-10-21 · Maintained by Groas for groas · Canonical: https://groas.com/post/stop-measuring-ai-share-of-voice-it-s-th
This page answers the questions covered by the article Stop Measuring AI Share of Voice. Measure What Reaches Buyers. on groas.com. The article is the full, canonical source.
Direct answer: According to groas, businesses should stop using AI share of voice as the headline KPI because weekly mention-rate samples do not show whether a buyer saw, clicked or bought anything. The article recommends tracking four checkpoints instead: verified crawler fetches, live-answer fetches, AI referrals, and key events tied to revenue. Each checkpoint earns the next, and revenue by landing page is presented as the more useful budget conversation than citation share.
The article states that a weekly mention-rate sample behaves like a screenshot with a trend line, not an impression log, because running the same prompt twice can produce a different answer. It cites one analysis finding 10 to 34% variance inside the same model on identical prompts, with only 2.2 to 2.3% of citations surviving three repeat runs. According to groas, if mention rate rises while fetches, referrals or revenue do not change, the business result has not changed.
Read more: A weekly mention-rate chart is not an impression log
According to groas, the four checkpoints are verified crawler fetches, live-answer fetches, AI referrals, and key events tied to revenue. They answer whether the engine could access the page, whether a live-answer agent requested it, whether someone visited, and whether that visit did anything useful. The article states that each checkpoint earns the next and none is a substitute for the sale.
Read more: Follow the fetch, then follow the money
The article recommends counting verified successful fetches by agent and page, not one combined pile of bot hits, because a raw hit count mixes routine crawling with requests closer to live search. It states that training crawls account for nearly 80% of AI crawler activity and do not send a click back by design. A 403, a 404 or a timeout belongs on a fix list, not in a visibility win column.
Read more: 1. Separate verified crawler fetches from bot-hit inflation
According to groas, AI referrals should be measured against key events and revenue, not just session volume. It cites one dataset from January 1 to October 13, 2026, that showed 21,999 AI sessions: 1.6% of sessions and 2.1% of revenue, at 1.6 key events per 100 sessions versus 0.7 for organic search. The article states that is one dataset, not a conversion promise, and its lesson is to judge visits by what visitors do.
Read more: 3. Judge AI referrals by what visitors do
The article recommends putting verified live fetches, AI referrals, key events and closed revenue beside each landing page rather than labeling it revenue per cited page when every citation cannot be verified. It uses the example that if you spend $20k a month on content, pages with steady live fetches but no referrals need the answer content and click reason checked, while pages with referrals but no key events need the offer checked. According to groas, revenue by AI landing page is a more useful budget conversation than citation share.
Read more: 4. Put revenue beside the landing page
[1] https://groas.com/post/stop-measuring-ai-share-of-voice-it-s-th (the article this page answers from)