AI Visibility Score

The AI Visibility Score is a 0–100 estimate of structural readiness — how prepared your site is for assistants to understand, retrieve, and recommend your brand. It is evidence-linked to crawled pages and technical checks. It is not a guarantee that a model will mention you tomorrow, and it does not read private model weights.

Score bands

How the score is composed

Each completed scan crawls your site (up to the crawl budget), runs a technical audit, and analyzes the corpus for recommendability. The overall score blends:

• Crawl & access — can retrieval bots reach key pages? Are robots.txt, sitemap, and rendering healthy?
• Entity clarity — consistent name, category, product, ICP, and trust signals across pages and schema.
• Answer coverage — FAQ, docs, comparison, and pricing surfaces that map to buyer prompts.
• Evidence strength — findings cite real URLs and excerpts from your crawl, not generic advice.

Related scores in the report

The report also surfaces a separate technical audit score (passed checks / total checks). Treat the AI Visibility Score as the headline readiness metric and the technical score as the infrastructure checklist underneath it. See Technical audit.

How to read your score in the product

1. 1

   Open the project report

   From AI Visibility (/app/p/[projectId]/visibility) or a public link at /r/[slug], note the ring score and band label.

2. 2

   Read the summary

   The 2–4 sentence summary explains how AI currently understands the business — product, ICP, and positioning.

3. 3

   Cross-check findings

   Strengths, weaknesses, and recommendations are evidence-linked. A high score with critical technical issues usually means content is strong but crawlers are blocked — fix both.

4. 4

   Plan the next lift

   Use the Weekly Execution Plan to ship the highest-impact fixes, then rescan to verify the delta.

How to raise the score

1. Allow retrieval crawlers (especially OAI-SearchBot, Claude-SearchBot, PerplexityBot) and decide separately on training bots like GPTBot — see Technical audit.
2. Publish clear Organization / Product facts in HTML and JSON-LD; keep naming consistent with your About and pricing pages.
3. Add answer surfaces: FAQ, docs, comparison, and alternatives pages that match buyer prompts.
4. Generate and ship assets from Content Studio, then mark tasks done in the weekly plan.
5. Install `vm.js` so verification ties to revenue, not only score movement.

FAQ