Scans

A scan is a single run of your Prompt set against your plan's AI platforms. Every scan produces a complete snapshot that feeds Visibility Score, Brand Coverage, AI Mentions, Market Position, Share of Voice, and the URLs tab inside Citations. See Concepts: Scans for the mechanics.

Triggering scans

Scans run on a schedule set by your plan; manual triggering is not exposed in the dashboard. Daily Monitoring runs every day and the Full Benchmark runs on the cadence defined for your tier.

On-demand scan triggering ships separately as part of the Console product.

Scan history

Every completed scan is retained. The per-domain scan history page lists scans with their completion time, Prompt count, and platforms tested. Clicking any scan switches the dashboard into "time-machine" mode so you can see exactly what it looked like at that moment.

The API equivalent is GET /domains/{domain_id}/scans, which paginates scan records. Pair it with GET /domains/{domain_id}/analytics?scan_id=... to pull the analytics payload for any historical scan.

Rescans

If a scan fails (for example, because an upstream AI platform returned errors on a large fraction of Prompts), SurfacedBy automatically retries within the retry budget defined for that scan. If retries do not succeed, the scan is marked failed and the dashboard reports the state. Your next scheduled scan will fill in the gap.

A manual rescan from the dashboard is not a customer-facing feature and will arrive with on-demand scan triggering.

Reading scan status

Scan state transitions are tracked strictly: queued, running, completed, or failed. A scan that appears "running" in the dashboard for a long time is still making progress; the progress bar updates as individual Prompts complete. If a scan has genuinely stalled, the system recovers it automatically and a follow-up scan runs.