> CiteGraph

CiteGraph measures which products AI assistants actually recommend. It asks the real questions buyers type into AI, like "best X for a small team" or "cheaper alternative to Y", across four engines (the ChatGPT family, Claude, Gemini, and Perplexity), repeatedly, and reports how often a brand is named, how often its own site is cited, and who wins the answers it loses. Every rate is published with a margin of error, because AI engines answer differently run to run and single-run scores are noise.

Each scan produces a shareable report: the verbatim answers, the competitors named instead, the pages the engines cited, and, for every lost question, the specific missing sentence that would put the brand in the answer, with drafted fixes ready to edit. Weekly re-scans track movement, flag question-level wins and losses, and surface newly discovered rivals and how AI frames each brand.

CiteGraph also publishes free public leaderboards of what AI recommends per category, re-run periodically with confidence intervals, including its own category, where it currently doesn't rank. Rank is measured, never sold. Plans start at $49/month with a 7-day trial. The leaderboards, research, and methodology require no signup.