How AI-visibility scoring works

LumaRank produces two headline scores (an AI Visibility score and an SEO Health score) plus the signals behind them. Here’s what goes into each.

AI Visibility

The visibility score (0–100) is a weighted blend of four measures:

• Mention rate: how often your brand appears in answers to unbranded prompts.
• Average position: where you land when models list options (being #1 beats being #5).
• Brand understanding: an LLM judge scores how accurately models describe what you do.
• Authority: whether models treat you as a credible, leading option.

We also break results down per provider (OpenAI, Anthropic, Gemini), track sentiment, and compare you against the competitors you configured.

Off-site presence

Models answer without browsing, so your visibility is a consequence of where you already show up. LumaRank measures the cause: mentions on Reddit, Wikipedia, the open web and Common Crawl, plus your share of voice versus competitors on each source.

AI signals

Ten technical signals tell us how easy your site is for AI crawlers to read: llms.txt, ai.txt, robots.txt, a sitemap, JSON-LD structured data, OpenGraph tags, an RSS feed, documentation, API docs, and a help center.

SEO Health

The SEO score is a weighted average across categories like crawlability, indexability, metadata, headings, content quality, images, internal linking, structured data, OpenGraph, mobile, performance and outbound links. Each category lists concrete issues with how to fix them, ranked so you can start with the highest-impact problems first.

How the two scores relate

AI Visibility and SEO Health are separate numbers, but they reinforce each other. Strong technical SEO makes your pages legible to the crawlers that feed AI models, and a healthy off-site presence is what pushes your brand into the answers in the first place. LumaRank surfaces the correlation explicitly: if your SEO is strong but AI visibility is weak, the action plan points you at presence and brand-clarity work; if both are weak, it starts with the fundamentals.

Tracking over time

A single audit is a snapshot. Because AI models refresh their knowledge and your presence shifts, the real signal is the trend. Re-run audits on a regular cadence (and after you ship a meaningful change) and LumaRank charts how each score and metric moves, so you can tell whether your work is actually paying off.

Want to act on this? See adding llms.txt and ai.txt.