Type a buying question into Perplexity and it answers with three or four names and a row of citations. There is no page two. If your project is not in the answer, you were never in the conversation. Here is how to get in, ranked by what 1,000 measured AI answers actually showed.
Perplexity does not rank pages. It composes an answer from sources it already trusts, then names the brands that recur across them. It retrieves live web sources and shows its citations, which makes it the easiest engine for a smaller crypto brand to break into. Get cited in the sources it reads, and you get named.
Three outcomes get confused as "ranking": named (your project appears in the answer text), cited (your content is one of the sources behind the answer), and absent (neither, where most crypto projects live).
Being cited is upstream of being named: engines name the brands that keep appearing in the sources they trust. And because answers are non-deterministic, a single screenshot proves nothing. Measure across many phrasings and many runs. That is the same reason every AI visibility tool gives a different score.
Perplexity is a retrieval engine first: it searches the live web, picks sources it trusts, and composes an answer with citations shown. That has three practical consequences we see in audit data:
ChatGPT blends browsing with training data; Gemini and Google AI Mode lean on Google's index. The moves below work across all of them, which matters because our study found the same shortlist but different winners on every engine. Optimizing for one engine leaves you invisible on the rest.
When we aggregated 1,000 answers, comparison and review sites (CoinBureau, BitDegree, Koinly, CoinMarketCap) filled the citation table. Engines answer "best X" questions with sources that already compare X. Absent from those pages, absent from the answer.
YouTube was the single most-cited source in our dataset: 319 citations, more than any website. Reviews, comparisons and explainers about your project are retrievable evidence. If nobody makes them, make them.
The only brand domains that cracked our top-10 sources run large educational hubs; kraken.com earned 246 citations with answer-pages, and Kraken tops three of five engines. Landing pages pitch; engines cite pages that answer.
Media citations (Forbes, crypto-native outlets) carry the trust layer. This is PR, but pointed: place where engines cite, not where impressions look good. We score 500+ crypto outlets on exactly this.
Reddit was fourth in our source table. Answers to real buyer questions, posted under your real name and title, become retrievable. Astroturf gets dissected in public, and engines read the dissection too.
"Best exchange for beginners" and "best no-KYC exchange" produce different names. Audit the phrasings your buyers use, find where you are absent, ship against the gap, re-measure on a cadence. That loop is the whole discipline.
Want to see which sources Perplexity cites for your category, and where you're absent?
Scan my brand free →Honest ranges from client work: Perplexity reflects new third-party coverage in days to weeks; ChatGPT and Gemini move slower where training data dominates. First measurable movement in about six weeks; a durable presence is a six-month build. Anyone promising overnight AI rankings is selling screenshots.
The engine-by-engine differences matter here. For the ChatGPT-specific version of this playbook, see how to rank your crypto project on ChatGPT.
Get named in the sources Perplexity already cites for your category: comparison articles, YouTube reviews, community threads, and your own question-answering content. Ask your buyers' questions on Perplexity, read the citations, and target those exact surfaces.
On Perplexity, yes, fastest of all engines. In our 1,000-observation exchange study it was the only engine that surfaced brands outside the consensus top tier, because it retrieves live niche sources rather than leaning on training data.
ChatGPT mixes training data with browsing, so it favors brands with long-standing presence; Perplexity favors current, retrievable coverage. Same playbook, different speed: expect Perplexity to move first and ChatGPT to follow. See how to rank on ChatGPT.
Run the same buyer-style prompts across all five engines, multiple times, and track how often you are named and cited. That is exactly what a citeOS audit does: 20 prompts across 5 engines, 100 observation points, free entry scan, weights published.
Growth marketer working in Web3 since its early days. Sagar has helped 75+ crypto and Web3 projects reach their target audiences, and runs Emergence Media. He built citeOS to make AI visibility measurable: 100-observation-point audits across 5 engines with every weight published in the methodology. Check your own citations with the free scan.
Five engines, twenty buyer prompts, 100 observation points. Published weights, reproducible citations, your baseline in about a minute. No card, no signup.
Scan my brand free →