# AI Visibility Reports

Interpret the current tracked-response visibility summary, its citation and sentiment context, and separate saved-report actions.

Canonical: https://docs.equalai.io/ai-visibility/ai-visibility-reports
Markdown: https://docs.equalai.io/ai-visibility/ai-visibility-reports.md

**AI Visibility Reports** summarizes analytics already available for the current project. Open it from the Optimization navigation or choose **AI Visibility Report** in **GEO Optimization**. Opening this page does not run prompts, query models again, or analyze your website's HTML.

## Read the visibility score

The score is the rounded percentage of responses in the loaded analytics set that contain a tracked brand mention. The denominator is responses, not unique prompts or competing brands. It is not share of voice and does not measure what fraction of all AI answers on the internet mention you.

| Score | Rating |
| --- | --- |
| 70–100 | Strong |
| 50–69 | Average |
| 30–49 | Below Average |
| 0–29 | Weak |

When no responses are available, the calculated score is zero. Treat missing analytics as a data-availability issue rather than a measured market position. The report has no independent date or model selector; its summary uses the loaded project analytics, not filters you may have applied elsewhere.

## Review the sections

- **Known For** lists up to five distinct prompt topics, excluding General and Others. These are tracked-topic labels, not a model-derived explanation of your reputation.
- **Quick Wins**, **Key Strengths**, and **Improvement Opportunities** give reviewable guidance. Confirm that an item applies before changing your content.
- **Brand Reputation** summarizes sentiment among brand-mentioned responses. Unknown or missing sentiment is treated as Neutral for this summary; the most frequent category is shown.
- **Citation Examples** shows up to three leading citation hosts across the loaded citations. A host's displayed sentiment follows the overall brand sentiment, not a separate sentiment analysis of that website.
- **Sample Search Prompts** shows a limited selection of tracked prompts, not the complete prompt inventory. The summary uses up to seven prompt examples and four competitor entries.

The citation total uses owned-domain citations when any are available; otherwise it falls back to all loaded citations. Do not assume every displayed citation count or example refers to your website. Inspect individual response and citation evidence before making that claim.

## Decide what to do next

1. Confirm that the project has tracked prompts and completed response data.
2. Read the score with its response coverage in mind. Changing the set of tracked prompts can change the result even if your website is unchanged.
3. Investigate representative prompts and cited sources rather than relying only on the headline score.
4. Use [Prompt Explorer](https://docs.equalai.io/ai-visibility/prompt-explorer) to improve prompt coverage or [Optimization](https://docs.equalai.io/ai-visibility/optimization) to investigate technical website signals.

The assembled report time is not proof that a new model run occurred. To obtain new responses, use your prompt-tracking workflow rather than reopening this page.

## Saved Report History, Share, and Export

The page can also show **Saved Report History**. Its **Share** and **Export** actions depend on an available saved report and target that saved report, not a newly captured copy of the current live summary. They are disabled when no saved report is available. A saved artifact can therefore differ from the summary currently displayed.

If you see **No analytics yet**, first establish tracked-response data for the selected project. Do not run website checks expecting them to populate this response-based summary.