We read your homepage and a sample of your inner pages, then generate six problem-led queries a real user might type into an AI chatbot — anchored to the topics your content actually covers. We’re simulating plausible queries, not pulling from real search logs. Snapshot at the time of the audit, AI answers shift over time.
How it works
Four steps from brand input to citation map
Read your homepage
We fetch your domain and extract what you actually do, so the queries match your real category — not whatever your brand name might suggest.
Generate 6 queries
Claude generates realistic queries a user might type into Perplexity, ChatGPT, or Gemini when researching your category. Brand name deliberately excluded so the audit is unbiased.
Perplexity runs all 6
Sonar API in parallel. We capture the answer text plus the citations array for each query.
Map your visibility
For each query: were you named in the answer? Was your domain cited? If not, which brands and domains were cited instead?