Scans

A scan is a single pass of your Prompt set against the AI platforms on your plan. It is the event that produces every metric in SurfacedBy: the dashboard, competitor view, and Citations all reflect the state of the most recent scan by default.

Two measurement scopes

SurfacedBy runs two scopes of scan:

• Full Benchmark - the complete tracked Prompt set against every AI platform on your plan. This is the official baseline used for Market Position, reports, and durable competitor comparisons. It refreshes Visibility Score, Brand Coverage, AI Mentions, Market Position, Share of Voice, and the URLs tab inside Citations.
• Daily Monitoring - a focused daily check so you catch movement between Full Benchmarks. Daily Monitoring updates the same metrics for the cells it covers and feeds the alert engine.

Both scopes flow into one composite read model: each (country, Prompt, platform) cell shows its latest known value, and the Full Benchmark provides the baseline that Daily Monitoring overlays for the cells it touches. See Measurement scope for the full read model and when each scope is the right source.

Cadence by tier

Cadence is set by your plan. Every paid plan runs Daily Monitoring every day plus a Full Benchmark on a plan-defined schedule. What differs between tiers is the Prompt budget and the set of AI platforms each scan runs against. See Plan limits for the current table.

What gets updated

A Full Benchmark refreshes every surface in the product:

• Visibility Score
• Brand Coverage
• AI Mentions (per platform)
• Market Position versus competitors
• Share of Voice
• The URLs tab inside Citations
• Suggested Prompts

Daily Monitoring updates the cells it covers for the same metrics and feeds the alert engine.

Comparison and mixed scope

When a delta or chart compares values across scope boundaries (for example, the most recent Daily Monitoring point against a Full Benchmark baseline), the dashboard renders a "Mixed scope" chip so the comparison is read as directional, not authoritative. See Measurement scope for the rules.

Triggering scans

Scans are scheduled, not on-demand. The dashboard and the public API do not expose a "run a scan now" button or endpoint. On-demand scan triggering ships as part of the Console product.

Scan history

Every scan is stored. You can walk back through past scans in the dashboard, and the API exposes both a scan list endpoint (GET /domains/{domain_id}/scans) and a scan-level analytics view (GET /domains/{domain_id}/analytics?scan_id=...) for time-machine analysis.