Assessment
Your paid diagnostic lives here.
The Assessment page shows the benchmark Visibility Score from your paid diagnostic engagement, plus how the score evolves over time week by week, with the driver behind each change.
You don't have a paid diagnostic yet. Choose what to do next:
The two scores are separate measurements taken on different dates. Compare each with its own history, not with each other.
VSS & LVS trend vs benchmark
The VSS reads your site and its sources; the LVS reads sampled AI answers. They measure different things, so the distance between the two lines does not show that AI has yet to notice a change. No fixed delay between a site change and AI answers has been established; each change is checked with a before and after comparison of the same questions.
What Changed Since First Benchmark
Two scores, two frameworks. The VSS section covers your paid-diagnostic infrastructure (the layers AI needs to find you). The LVS section covers live AI behavior (what AI is actually saying right now).
Visibility Stack Score
Paid-diagnostic infrastructure metrics — what AI needs to find you. Frozen until your next paid audit.
| Layer | Weight | First Benchmark → Current | First Benchmark | Current | Change vs First | Change vs Previous |
|---|
Visibility Stack Layers (click any card for layer description)
Live Visibility Score
Live behavioral metrics — what AI is actually saying about you. Updated on every prompt re-test (typically weekly).
| Component | Weight | First Benchmark → Latest Run | First Benchmark | Latest Run | Change vs First | Change vs Previous |
|---|
7-Component Detail
Each component answers a distinct question from row-level evidence. Click any card to drill into the underlying mentions, themes, or competitor data. No metric repeats another. Computed from the last 4 measurement cycles of prompt × platform results, one completed core audit per cycle, so the number moves when a cycle lands and at no other time. The four cycles ask the same questions on the same tools, so they are not four independent samples: pooling them limits the effect of one unusual cycle without removing it.