Reading Your Results
How to interpret everything AIVIS.space shows you after a scan.
Start with the Overview
After a scan completes, check your Overview first. The five KPI cards at the top give you an at-a-glance read:
- Visibility Score — Is AI mentioning you? A low score means you're not in the conversation.
- Integrity Score — When AI mentions you, is it accurate? A low score means AI is getting your facts wrong.
- Weekly Momentum — Are things getting better or worse compared to last week?
- Competitive Position — Where do you rank relative to your tracked competitors?
- Risk Radar — Are there active inaccuracies you need to fix?
If the Overview looks healthy, you may not need to dig further. If any of these flags a problem, go to the relevant section to investigate.
Go to Scan Summary for this scan's results
Scan Summary shows the most recent scan results. Each row is a specific query × model combination.
What each row tells you:
- The query that was scanned
- Which AI model responded
- The verdict (Accurate, Hallucinated, Mixed signals, Competitor bias, Unverified, Insufficient evidence)
- The Integrity Score for that specific response
Sort by Integrity Score (lowest first) to see where the biggest problems are.
Understanding the verdict
The verdict is the most important field in each row. It tells you how AI responded to that query.
| Verdict | What it means |
|---|---|
| Accurate | AI described your brand correctly |
| Hallucinated | AI stated something that contradicts your truth record |
| Mixed signals | AI got some things right and some things wrong |
| Competitor bias | AI recommended a competitor in response to a brand-relevant query |
| Unverified | AI made claims that can't be checked against your truth record |
| Insufficient evidence | AI didn't say enough about your brand to score |
See Verdicts for detailed explanations of each one.
Open the Audit Ledger for the details
The Response Audit Ledger is available on Growth and above.
When you see a Hallucinated or Mixed signals verdict and want to understand exactly what went wrong, click into it to open the Audit Ledger entry for that scan.
Left panel — what AI said: The full AI response with color-coded highlights.
- Green = AIVIS.space confirmed this matches your truth record
- Red = AIVIS.space found a contradiction with your truth record
- Orange = AIVIS.space couldn't verify this against your truth record
Right panel — the evaluation:
- Verdict tab — The overall verdict and score explanation
- Truth tab — Side-by-side comparison of what AI said vs. what your truth record says
- Citations tab — Which URLs appeared in the AI response
- Provenance tab (scan details — model version, timestamp, and scan ID) — Metadata about when and how the scan ran
Read the Truth tab to understand exactly which facts AIVIS.space checked and what it found.
Check your Actions
New Actions appear automatically after each scan. Go to Actions and filter by New to see what AIVIS.space found in this scan.
Work Critical and High priority actions first. They represent inaccuracies that appear most frequently or have the highest impact on your score.
For each action:
- Read the action card — it shows the exact AI quote and the source of the conflict
- Decide whether to fix it, dismiss it, or snooze it
- For Accuracy Fix and Hallucination Risk actions: fix the issue on your website, then mark as Actioned
- AIVIS.space automatically creates a Verify Fix action 48 hours later to confirm the fix worked
Check per-model performance
Go to AI Model Visibility to see how each AI model is performing separately.
If one model (for example, Gemini) consistently produces lower Integrity Scores than the others, that model is getting your facts wrong more often. The actions that come from that model's scans are where to focus first.
Track your trend, not just today's result
A single scan result can be misleading. An Accurate verdict today doesn't mean everything is fixed — the next scan may show a Hallucinated verdict because AI updated its training data.
Use the weekly momentum and trend lines in Overview to track whether your score is improving over weeks. That trend is what matters for your brand's AI health.