How Do You Get Cited by ChatGPT, Claude, Gemini, and Perplexity?
There is no single way to get cited by an AI engine, because the four major engines do not select or cite sources the same way. Only about 11% of source domains get cited by both ChatGPT and Perplexity for the same query. Optimizing for one engine can do nothing at all on another. Searches for "how to get cited by ChatGPT" specifically are up 247% month over month as of this week, which suggests more buyers are starting to ask this question per engine rather than assuming one answer covers all of them.
Four engines, four playbooks
ChatGPT trains on data first and searches live only on demand, citing in roughly 2.8% of answers as of last measurement, up from under 1% a year earlier. When it does cite, Wikipedia dominates, followed by Reddit, Forbes, and G2. Since around May 2026 it has started surfacing clickable brand links inside answers, which makes it a referral channel now, not only an answer box.
Perplexity retrieves and reranks live pages at query time, and Reddit dominates its citations. It carries a strong recency bias: content published in the last 30 days gets cited roughly three times more often than older content. It is also the cleanest engine to measure, because it reads without a login wall.
Gemini grounds every answer in live Google Search and decomposes a prompt into sub-queries before it searches, a behavior called query fan-out. It has no separate crawler of its own; it uses Googlebot plus the Google-Extended opt-out flag, and it carries the same recency bias Perplexity does.
Claude searches the web via Brave and cites less often than the others, more selectively, favoring depth, clear structure, and verifiable third-party corroboration over a brand's own self-description. A page written to sound authoritative about itself is exactly the kind of page Claude tends to skip.
What this means in practice
Pick the engines your buyers actually use before optimizing for any of them. A B2B SaaS company whose buyers live on Perplexity for research should build for recency and Reddit-adjacent presence first. A consumer brand whose buyers ask ChatGPT casual questions should worry more about Wikipedia and G2 than about Reddit. Nobody gets to skip the work of finding out which engine matters to their own buyer before spending on any of this.
Final thoughts
We track our own citation rate across all four engines on a fixed schedule specifically because a single engine's result would tell us almost nothing about the other three. If a vendor pitches you one tactic that works "for AI search," ask which engine they measured it on, and ask whether they checked the other three.
Regards,
Charles Stack & Sela, AI Optimization — Coworkers.Global
Cite this: "Only about 11% of source domains are cited by both ChatGPT and Perplexity for the same query; each of the four major AI engines selects and cites sources by a materially different mechanism." Per-engine facts last verified August 2026. Search-volume figures via DataForSEO, captured September 14, 2026.
"FAQtopus" is a trademark of Coworkers Global, Inc.
Comments