Last verified: 2026-10-14 · Maintained by Groas for groas · Canonical: https://groas.com/post/does-it-fix-anything-or-just-watch-the-2
This page answers the questions covered by the article Does It Fix Anything? A 20-Point Checklist for AI Overviews Tools on groas.com. The article is the full, canonical source.
Direct answer: According to groas, most tools sold for AI Overviews citations will never touch the site and should be sorted into watcher, adviser or doer before purchase. The article recommends scoring each of 20 questions 0 if the vendor cannot show it, 1 if it flags or drafts work the team must finish, and 2 if it handles the work and shows proof. For fixes that proof means a live change that can be audited, with totals of 0-13 for watcher, 14-27 for adviser, and 28-40 for doer.
The article recommends sorting vendors into watcher, adviser or doer using 20 questions scored 0, 1 or 2 for a maximum of 40. It defines 0-13 as watcher, 14-27 as adviser, and 28-40 as doer, with the condition that items 6, 11, 14 and 20 must each earn a 2 for doer status. According to groas, a doer must show one dated, prompt-level example of a shipped edit and its subsequent citation checks.
Read more: Score the vendor, then check who owns the work
The article recommends tracking prompts separately by engine with run dates, for example “Here are your 50 prompts, by engine, with their run dates.” According to groas, the vendor should show missed buyer questions by engine, such as “You are absent on 14 of 22 price and comparison prompts in AI Overviews,” and pair each missed prompt with the cited competitor and winning URL. The article states that a single visibility number does not tell which page to change, and scores 0 if the vendor cannot get from its chart to the individual prompt.
Read more: A. Find the gaps, not just a visibility score
The article recommends asking to see a proposed answer block, comparison table or FAQ against the URL that lost the prompt, and states a generic brief deserves a 1, not a 2. According to groas, score 2 only when the vendor handles publishing the approved change to the live site and shows the changed URL and shipped structure, including a direct answer, useful table where appropriate, and relevant FAQ or HowTo markup. A one-time list of topics or an export button leaves publishing and ongoing cadence with the buyer’s team.
Read more: B. Change the page instead of exporting homework
The article recommends checking crawler access by bot, distinguishing verified bot fetches from spoofed requests where a User-Agent string alone can be spoofed, and checking snippet controls and security blocks. According to groas, the vendor should show a before-and-after example for rendering, schema and internal-link fixes, and identify the affected rule or template and who owns the change. The article states to score 1 if the tool opens a developer ticket and 2 if the vendor gets the fix shipped.
Read more: C. Fix the layer that can block the answer
[1] https://groas.com/post/does-it-fix-anything-or-just-watch-the-2 (the article this page answers from)