AI Visibility

Why this exists

People increasingly ask an assistant before they open Google. The assistant reads a handful of pages and cites a few of them. Your ranking position doesn’t decide which ones — parseability, attribution and sourcing do.

You can hold position one and still be absent from the answer people actually read. This screen tells you whether that’s happening.

Brand presence vs link presence

• Brand presence — the share of sampled prompts where your brand is mentioned by name, whether or not a link is given.
• Link presence — the share where your domain is actually cited as a source.

Link presence is almost always the smaller number, and the gap is diagnostic. Being named but not cited means the model knows who you are but doesn’t consider your page the place to send someone — usually a structure problem, not an authority one.

Share of voice

The same two figures are calculated for competitors and shown side by side, so you can see who owns the answer space in your category rather than just tracking yourself in isolation.

Prompt discovery is region-aware across roughly thirty-seven regions, because assistants answer differently depending on where the question comes from.

Which engines

Five surfaces are covered: ChatGPT, Perplexity, Gemini, Google AI Overviews and AI Mode. They behave differently — Perplexity cites heavily, Overviews cite sparingly — so a single blended score would hide more than it showed.

Improving your position

The eleven AI Readiness checks are the practical lever, and nine of them ship a fix:

• FAQ schema and direct-answer formatting, so a model can extract a clean answer
• Author attribution and cited sources, so it has reason to trust the page
• Supporting statistics and expert quotes, so there’s something specific to quote
• Content freshness signals and llms.txt, so crawlers know what’s current and what they may use

These aren’t tricks. They’re the same things that make a page genuinely useful to a human reader in a hurry.

What this can and can’t tell you

This is a sampled read of how you appear right now, not a continuous tracker with guaranteed coverage of every possible prompt. Assistants are non-deterministic — the same question can produce different sources on different days.

Treat the numbers as a directional signal across a set of prompts, not a precise ranking. The competitor comparison is more reliable than any single figure, because both sides are sampled the same way.

Next: PDF reports →