1) Run your prompts
Scheduled sampling collects answers, and with them every source the assistants cite.
When ChatGPT or Perplexity answers with citations, it is showing you exactly which sites shape your category. CitePrism collects those citations from every sampled answer and categorizes them: your site, competitors, review platforms, communities, institutions. That list is your GEO to-do list.
A visibility dip tells you something changed. The citation list tells you what to do about it, because it names the pages the assistants are reading.
If one review site keeps getting cited, your presence there matters more than another blog post.
When a rival comparison page starts appearing in citations, you will see it the week it happens.
Citation share shows whether your content is becoming a source the assistants rely on.
How it works
Scheduled sampling collects answers, and with them every source the assistants cite.
Citations are grouped by domain and category, ranked by how often they appear.
Fix your presence on the sources that matter, then watch citation share respond.
No. Perplexity cites on nearly every answer, ChatGPT cites when it searches the web, Gemini often grounds in Google results, and Claude cites when browsing. CitePrism captures citations wherever the engine provides them and tracks coverage per engine.
Each cited domain is classified as your own site, a competitor, a review platform, social or community content, search properties, institutional sources, or other. The categories turn a long URL list into a readable picture of who shapes your category.
Backlinks measure who links to you on the web. Citations measure which pages assistants actually pull into answers, which is a much smaller, more decisive set. A site with modest backlinks can dominate assistant citations in a niche, and you want to know when that site is not yours.