Measurement Scope

Every metric in SurfacedBy is produced by one of two measurement scopes: Daily Monitoring or Full Benchmark. Daily Monitoring is a focused daily check of a subset of your tracked prompts on a rotating pair of platforms. Full Benchmark checks your complete tracked prompt set across every tracked platform on a slower cadence. Full Benchmark is the official baseline used for rank, reports, and competitor comparisons.

When each scope runs

The cadence depends on your plan.

• Daily Monitoring runs every day on every paid plan.
• Full Benchmark runs on the schedule set by your subscription tier.

See the Plans page in the dashboard for the exact frequency on your tier.

How displayed values are composed

Think of a scoreboard with one cell for every (country, prompt, platform) combination tracked on your domain. Each cell holds the most recent answer the AI gave for that combination.

• A Full Benchmark scan refills every cell on the scoreboard.
• A Daily Monitoring scan rewrites a smaller subset of cells. Every other cell keeps its previous value.

The dashboard shows the scoreboard as it stands right now. No cell is ever blank between scans because the previous value carries forward until something newer replaces it.

A concrete example. Suppose Full Benchmark tested 100 prompts on Monday and Daily Monitoring rechecked 20 of them on Tuesday. The Tuesday values overwrite the Monday values for those 20 prompts. The other 80 still show Monday's Full Benchmark result. The Overview KPIs, AI Visibility cards, Citations charts, and Tracked Prompts table all read this composite.

Aggregate windows versus current state

The composite answers "what is the state right now". A different reader answers "what happened in the last N days".

• Composite is used for current-state values: Visibility Score, Market Position, the scoreboard, sparklines, per-source rank, and same-scope deltas. Each tuple collapses to its single most recent response.
• Window aggregate is used for counters: Brand Coverage percent, total AI Mentions in a window, "X of Y sources cite you", "Total citations", and CSV exports labelled as totals. Every response in the window contributes to the count.

If you would manually count "every AI answer in the period" to reproduce a number, that number comes from the window aggregate. If you would check "what does the latest scan say right now", that number comes from the composite.

Mixed-scope comparisons

A delta compares two snapshots of the same metric. By default the dashboard finds a previous snapshot of the same scope: a Daily Monitoring value compares to the previous Daily Monitoring value, and a Full Benchmark value compares to the previous Full Benchmark value.

When no same-scope predecessor exists, the dashboard falls back to the most recent prior snapshot of the other scope and labels the result with a "Mixed scope" chip. Mixed-scope comparisons are directional, not exact: the underlying scans tested different prompts and platforms, so the magnitude of the change is approximate. Read the chip as "the trend is going up or down" rather than as a precise number.

Market Position

Market Position is the rank value used in reports, the Competitors page, and the Overview KPI tile. It is anchored to Full Benchmark: the value is recomputed only when a Full Benchmark scan completes. On Daily Monitoring days, the dashboard shows the persisted Market Position from the most recent Full Benchmark. This avoids day-to-day rank flicker driven by partial measurement.

Recommendations work the same way. The opportunity list is produced only by Full Benchmark scans. Daily Monitoring does not add or remove opportunities.

What this means for you

• Treat the default dashboard values as the current state of the world. They are not the result of a single scan; they are the composite of every scan to date.
• Treat aggregate counters ("in the last 30 days", "total citations") as full-window counts, not as snapshots of the latest scan.
• Treat a "Mixed scope" chip as directional. The change is real; the exact size is approximate.
• For rank, reports, and competitor comparisons, the Full Benchmark anchor is the source of truth.