Score Movement
Your visibility score moved. Why? Score Movement is the dashboard panel that answers that question — by showing you exactly which prompts gained, lost or shifted rank since the previous scan.
Where to find it
Dashboard → Overview, second row of cards (spans two columns). The panel only appears when there's a previous scan to compare against — so you'll see it for the first time after your second scan.
Three categories of change
| Type | What it means | Score impact |
|---|---|---|
| Lost mention | You appeared in this prompt last scan but don't anymore. | Negative — proportional to whether you were also recommended. |
| New mention | You appeared in this prompt for the first time. | Positive — proportional to whether you were also recommended. |
| Rank changed | You appeared in both scans but your position shifted (e.g., #2 → #4). | Either direction; uses 1/rank scoring so position #1 → #2 hurts more than #5 → #6. |
How impact is calculated
Each score driver is annotated with an estimated point impact. The math:
- Lost or gained mention: ±30 / total responses (mention component)
- If the lost/gained mention was also recommended: ±30 / total responses (recommendation component)
- Rank shift: (1/new_rank − 1/old_rank) × 25 / (responses/3)
The top 5 movers (by absolute impact) are surfaced in the panel. Click See all changes → to jump to the AI Responses page filtered to lost mentions.
How to use it
-
Open Score Movement after every scan
It's the fastest way to know whether the score moved for a real reason or just noise.
-
Click a "Lost mention" row
You'll land on AI Responses filtered to that exact prompt — read the answer to understand why you fell out of the AI's vocabulary.
-
Look for patterns across drops
Multiple losses in one topic = positioning issue in that topic. Multiple losses on one engine = an engine-specific problem (Gemini retraining, for example).
-
Celebrate the gains
Gain rows tell you what's working. If you shipped content for a prompt and it now mentions you, you've closed the feedback loop — record the win for next quarter's review.
AI engines are non-deterministic. A single lost mention isn't a crisis — but the same prompt losing on every scan for three weeks running is a real positioning gap. Use Score Movement to spot patterns, not panic over snapshots.
What's NOT in Score Movement
- Sentiment changes (covered separately in the Sentiment Overview)
- New prompts you added between scans (no comparison baseline yet)
- Citation source changes (covered in Cited Pages & Domains)
Was this page helpful?
Thanks for the feedback — we read every response.