The LLM visibility tool that shows who AI recommends, and why

An LLM rank tracker measures your presence inside AI answers, not a position on a page of links. For each prompt and engine it records whether the answer names you, how it treats you, whether it cites your pages, and who else it recommends.

AI answers have no fixed rank, so the useful numbers are rates across many answers: how often you are mentioned, cited and recommended, and in what position among the brands named. LogNorm shows every rate with the number of answers behind it, so a 50% on two answers never looks like a 50% on two hundred.

LogNorm tracks ChatGPT, Gemini and Google AI Overviews, asked in the consumer products your buyers use rather than through developer APIs. The products search the web, pick sources and phrase answers differently from the raw models, so the API answer is not the one your buyer reads.

Every engine is asked from the country of your website's market, because answers change with location. For Google AI Overviews alone, see the AI Overview tracker.

Because they fail for different reasons and need different fixes. Merging them into one visibility score hides which problem you have.

• Mentioned, not cited: the engine knows you from third parties, so what it says about you is out of your hands. Make your own pages the evidence.
• Cited, not mentioned: your page informs the answer without earning the recommendation. It probably teaches without making the case for your product.
• Absent: look at who is named instead and which pages the engine cited for them.

LogNorm reports three rates from these facts: mention rate, citation rate and recommendation rate, where the answer makes you the top pick or one of its recommendations. Each mention also carries a type, from top pick to negative, with LogNorm's confidence.

Track your competitors through the same answers you are tracked in. LogNorm reads every answer for every competitor you follow and puts you side by side.

• A table with your mention, citation and recommendation rates next to each competitor's, the change since the previous run in percentage points, and the average position among brands named.
• Per competitor: their rates after each run, their strongest engine, and the prompts they win while you are absent or only an alternative.
• Their own pages that engines cite, which shows what kind of page earns the citation.
• Third-party pages that mention them and not you.
• Brands that keep appearing but aren't on your list, with LogNorm's call on whether each one competes with you.

Track as competitor adds a brand to your tracked competitors, so its website is analysed too.

A Google rank tracker reports one position per keyword; an AI rank tracker reports how often you appear across many answers that change from one ask to the next. You need both, and LogNorm keeps both on the same website.

The Sources tab lists every page the engines drew on, over the last four runs by default. LogNorm records three ways an engine uses a page: cited among the answer's sources, linked inline in the text, or searched while researching and then left out.

Domains are grouped by kind (yours, competitors', review sites, publishers, forums, docs, directories, video, social and marketplaces) with each kind's share of cited answers. Pages show their type, such as listicle, comparison or forum thread, and whether they mention you. This is where most visibility wins start: a listicle that every engine cites and that leaves you out is a clear outreach target.

LogNorm opens the most-cited third-party pages to check who they mention. Pages it couldn't open read Not inspected and are never counted as leaving you out.

The Historical runs tab shows each prompt day by day for the engine you pick. Each cell says whether the answer named you and how, and whether it cited your pages.

Show the last 7, 14, 30 or 90 days with answers, and click a cell to read that day's answer. Export CSV downloads one row per prompt, engine and day with whether you were named, how, your rank and whether you were cited. Runs happen on your schedule: daily, twice a week, weekly (the default) or monthly.

You turn it into work. After each run LogNorm writes learnings, each with its evidence and what to do, ranked by how much evidence supports them.

Get listed points at a cited page that lists competitors and not you. Pages to build names a page type competitors get cited for that you lack. Engine gaps, Lost ground and Competitors gaining catch movement between runs. The strongest learnings become moves on your Growth Plan on their own, and the Impact tab later shows whether the work you shipped changed the answers.