Perplexity's crawler — it fetches pages in real-time to back the citations Perplexity shows inline.
PerplexityBot runs two modes: a scheduled crawl of allowed sites (for retrieval) and a real-time `Perplexity-User` agent that fetches pages while a user is actively asking a question. The second mode is why a fresh answer appears in Perplexity within seconds of publishing.
Because Perplexity surfaces citations inline, the bot heavily weights pages with clear author bylines, dates, and canonical URLs. Ship E-E-A-T signals and Perplexity cites you by name.
OpenAI's web crawler that fetches pages for ChatGPT's retrieval and training data.
Anthropic's crawler that fetches pages for Claude's retrieval and training.
Experience, Expertise, Authoritativeness, Trustworthiness — Google's content-quality heuristic, also used by AI retrievers.
The percentage of AI responses about a topic that link to or name-check a specific source.