Visibility scans now record whether each AI platform actually searched the web when answering a prompt about your brand — and whether your site appeared in its citations. The Explorer shows account-level rates at a glance, and individual prompts display a pill showing how many engines searched live versus answered from built-in knowledge. This matters because an AI citing your site from a live search is a different signal than one recalling it from training data.
A new Fact Sheets page (Business Info → Fact Sheets) surfaces the structured documents your chat agent uses to ground its answers — your AI Visibility standing, Technical SEO readiness, Action Plan, and others. Each sheet shows when it was last refreshed, and you can force a manual refresh per sheet or all at once. Previously these were recomputed invisibly on every chat turn; now you can see exactly what the agent knows and how current it is.
The AI Visibility fact sheet now includes your overall score, per-platform recommendation scores, scan-over-scan trend, and topic/persona coverage — updated automatically when a scan completes. A timing issue where the fact sheet could show the previous scan's score right after a new scan finished (exactly when the "Scan Complete" notification lands) is now corrected.
The Business Review's "What we found" section is now structured into a TL;DR, a nuanced read, and a "What to do" — rather than a single block of text. The recommendations now match the mechanism: if your brand is already appearing in searched answers, you get third-party footprint advice rather than more on-site content edits.