AI Visibility Score

Overview

The AI Visibility Score (/visibility) is a single 0–100 metric that summarizes how well AI systems surface your brand when answering buyer questions in your category. Obsurfable derives the score from your latest monitored prompt run: each company prompt is evaluated for brand mentions, domain citations, top-entity listings, and competitor presence.

The report includes a trend chart, score breakdown by dimension, per-query-type comparison, competitor mentions, per-prompt analysis, and a prioritized action plan so you know what to fix first.

Who it's for

• Marketing and growth teams who need one number to track AEO progress over time
• Founders and product marketers validating whether content and positioning changes improve AI answers
• Agencies reporting visibility improvements to clients after site or content updates

Prerequisites

1. Complete onboarding and add your company profile.
2. Set up company prompts with buyer questions you want to monitor.
3. Create a prompt run and use Ask ChatGPT on at least 3 prompts to unlock the full visibility report.

Until you have enough results, /visibility shows a prompt to run your company prompts first.

Step-by-step

View your score

1. Open Visibility in the sidebar or go to /visibility.
2. If a report is ready, you'll see your current score on a gauge, a verdict summary, and when it was last updated.
3. Review the trend chart to see how your score has changed across runs (delta badge shows change since the previous run).

Understand the breakdown

1. Scan the stat cards: prompts that mentioned you, average citation score, competitors cited, and mention rate.
2. Open Score breakdown to see average points for mentions, citations, and top-entity listings.
3. Check Score by query type to compare category discovery, direct brand, and comparison prompts.

Find competitor gaps

1. Scroll to Competitor mentions to see which competitors appeared in AI responses.
2. Review Per-prompt analysis for each question—whether you were mentioned, your per-prompt score, and which competitors won instead.
3. Use Track competitors to manage your competitor list in Company settings.

Act on recommendations

1. Read the Action plan at the bottom of the report for prioritized fixes tied to your weakest prompts.
2. Follow links to the related prompt run or company prompts to address specific gaps.
3. After making site or content changes, create a new prompt run and compare results over time.

Tips

• Run prompts regularly (or enable scheduled runs on Pro) so the trend chart reflects real changes, not one-off snapshots.
• Add competitors in Company settings—competitor gap analysis is more useful when Obsurfable knows who to watch for.
• Focus on discovery prompts where you aren't mentioned; these represent category questions buyers ask before they know your brand.
• Pair the visibility score with Insights (Pro) for detailed AEO recommendations on the same run.

FAQ

Why don't I see a score yet?

You need at least one prompt run with Ask ChatGPT results on 3 or more prompts. If you have no runs at all, start from Prompts and create a new run.

How is the score calculated?

Each prompt is scored on mentions, citations, and whether you're listed among top entities in the AI response. Obsurfable averages these across all evaluated prompts in your latest run to produce the 0–100 composite score.

Does the score update automatically?

The score reflects your latest completed prompt run. Create a new run and re-ask prompts to refresh the report. On Pro, scheduled prompt runs can keep this updated automatically.

What's the difference between mention rate and citation rate?

Mention rate is how often the AI names your brand. Citation rate is how often it links to or cites your domain. You can be mentioned without being cited—and citations are often stronger signals for retrieval.

Can I see the score for an older run?

The main /visibility page shows the latest run. Open a specific prompt run to see score impact for that run, or use Compare runs to see mention and citation changes between two runs.