How often the answer names you at all

No. There is no position one inside a paragraph. Visibility is the share of answers on your tracked prompts that named you at all — a hundred answers, and the score is how many of them mentioned your brand.

Two things, and the panel separates them. A real change — a page the engines started or stopped trusting — and ordinary variance, because engines are not deterministic. The confidence interval and the run count sit next to the score so you can tell which one you are looking at.

AI visibility is how often, in which position and with what tone a brand appears in the answers generated by assistants such as ChatGPT, Gemini, Perplexity and Google AI Overviews. It is not a search ranking: a brand is either named inside the answer text or it is absent from the decision entirely. Rixe measures all three dimensions daily and reports them as one visibility score per category.

An SEO tool measures where a page ranks on a results page. Rixe measures whether a brand is named inside an AI-generated answer, which position it holds among the brands mentioned, and which sources the model used to build that answer. The two rarely agree: a page can rank first in search and never appear in the AI answer to the same question, because the model draws on third-party sources such as review sites, forum threads and comparison articles.

Rixe runs a defined set of category prompts through each engine's live interface once a day and parses the answer that comes back. For every run it records whether the brand was named, its rank among the brands mentioned, the sentiment of the sentence around the name, and every source the answer cited. No score is published from a single run: below five runs a prompt is marked insufficient sample, and every score ships with a Wilson confidence interval.

Rixe tracks ChatGPT (GPT-5 Search), Gemini, Perplexity, Google AI Overviews, Google AI Mode, Microsoft Copilot, Yandex Neuro, Grok, Claude Sonnet 4, DeepSeek, Qwen and Mistral. Every plan covers every engine — the limit is the number of prompts tracked, never the number of engines.