Understanding your AI Visibility Score

The AI Visibility Score is a 0–100 summary of how accessible, readable, and trustworthy your page is to AI systems. Here is what it measures and what to do with it.

What the score is

The AI Visibility Score is a single number from 0 to 100. It rolls up the main signals an AI system needs to reliably read, understand, and use a page: whether bots can reach it, whether the important content is in the HTML response, whether the page has enough trust signals to be attributed, whether the server responds fast enough, and — once the page’s content has been analysed — whether the page actually answers the questions AI asks about its topic.

The score is accompanied by a status label, which reflects how urgent the findings are. The status matters more than the raw number — and the two are tied together, so the number can never read better than the findings behind it.

The five things it measures

1. AI Crawlability

Can the major AI bots access this page at all? If robots.txt blocks a bot, or if the live request fails with a 4xx/5xx, the page is unreachable regardless of how good the content is.

2. Content visibility

Is the important content in the raw HTML response? Pages that deliver their content only after JavaScript runs are at a disadvantage because many AI systems fetch the raw response rather than waiting for a browser-rendered version.

3. Authority

Does the page give AI systems enough context to understand what it is and trust it as a source? This includes structured data (JSON-LD), clear page-type classification, author metadata, and publication dates.

4. Speed and stability

Does the server respond fast enough, and is the page stable while it loads? TTFB (time to first byte) is the most important metric. Slow pages are harder to include reliably in AI retrieval pipelines. These metrics are surfaced inside the AI Crawlability tab (TTFB) and Content Visibility tab (CLS/INP) rather than a separate tab.

5. Answer Readiness

Once the page’s content has been analysed, does it explain its topic completely enough for an AI system to build an answer from? Where Content Visibility asks can AI read the page?, Answer Readiness asks can the page answer the questions AI is asking? — the questions a page on this topic should answer, and the concepts it should integrate. It’s produced by the Content Coverage analysis, appears once a page has been analysed, and blends into the overall score once enough of your pages are covered.

Status labels

The label comes from the score:

Score Status What it means
85–100 Excellent No critical blockers detected. The page meets baseline requirements for AI retrieval.
70–84 Needs work One or more moderate issues found. The page is reachable but something is limiting its visibility or trustworthiness.
55–69 Concerning A hard blocker is present, or several moderate issues are stacking up.
40–54 At Risk Multiple problems across more than one dimension.
0–39 Critical The page fails at a basic level — usually it cannot be reached or cannot be read at all.

The findings cap the number. A page with a critical finding cannot score above 69, and a page with a warning finding cannot score above 84 — however well the other dimensions do. This is deliberate: it stops a page with a hard blocker from showing a reassuring number. It also means a verdict reading Critical beside a score labelled Concerning is the cap working, not two parts of the report contradicting each other.

When there is a hard blocker, fix it before anything else. Lower-priority issues will not help while it is unresolved — and the score cannot rise past the cap until it is cleared.

How to use the score

Start with the status, not the number. A Critical status always means there is something to fix urgently. An Excellent status means the baseline is healthy — you can then focus on deeper improvements like Authority, Prompt Fit, and content quality.

Use the dimension bars to find what’s dragging the score down. The score card on the Action Plan breaks into AI Crawlability, Content Visibility, Authority, Speed & stability, and an Answer Readiness bar once the page’s content has been analysed. Each links to the tab behind it. Speed metrics don’t get their own tab — TTFB appears inside AI Crawlability and CLS/INP inside Content Visibility. If one bar is short and red, start there.

Work the plan from the top. The Action Plan already orders the findings for you, so you don’t have to decide what to tackle first.

Let Optimizations write the content fixes. For the content and structure findings — Content Coverage, Prompt Fit, Prompt Discovery, and Authority schema — you don’t have to draft the change yourself. On any paid plan, Optimizations writes the fix from your page, ready to paste in and re-audit.

What the score does not tell you

A high score does not mean your page will be cited by AI. The score measures technical access and structural quality. It does not predict whether your content is the best answer to any specific question. A page that scores 90 but gives a vague or off-topic answer will still be skipped.

A low score does not mean the page is broken. A low score can reflect an intentional configuration choice — like blocking certain bots on purpose, or a product page that does not need structured article data. Use the detail tabs to understand the actual findings.

The score does not update automatically. It reflects the page’s state at the time of the audit. Run a new audit after making changes to measure the impact.

Next steps