How we test what AI says about you

Every number we show comes from an answer you can read.

The questions

We write questions the way customers type them, for example “Who is the best family law attorney in Plano, TX?”. The free check uses four of them plus one about your business by name. Subscribers get 25 to 50 tracked questions, grouped into topics.

The assistants

ChatGPT is tested through OpenAI’s model with live web search, and Gemini through Google’s model with Google Search grounding. The free check uses these two. Weekly tracking on a plan also asks Perplexity, through its Sonar search model. It also reads Google’s AI Overview for each question from the live US Google results; when Google shows no Overview for a search, that counts as not named.

Asking more than once

AI answers change from one ask to the next, so the free check asks each question twice per assistant. Weekly tracking trends across many answers, which smooths out the noise.

The score

The visibility score (0–100) counts how often you are named in answers to discovery questions and how high in the answer you appear: first place counts most. Questions that name your business are tracked for accuracy but left out of the score. The free check’s readiness score averages five website checks: AI crawler access, structured data, answer-ready content, entity clarity and third-party footprint.

What we store

The full answer text, the sources the assistant cited, the date and the assistant.

Our crawler

When you run a check we fetch your homepage, robots.txt, llms.txt and sitemap once, identifying as boostmyaisearch-audit.