How do I get cited by Perplexity?

How does Perplexity work?

Perplexity combines its own freshness-focused search index with live retrieval: for every question, the system fetches candidate pages, extracts the relevant passages and builds an answer in which every statement is mapped to a numbered source. That sets Perplexity apart from classic chatbots: there is no answer without a source list. The platform has been a citation machine from day one — around 780M queries per month in May 2025, growing more than 20% month over month according to the company (CEO Aravind Srinivas, June 2025). Perplexity has not published a more recent official monthly figure.

How does Perplexity pick its sources?

What matters is the passage, not the brand size: for each claim, Perplexity looks for the source that supports it most directly. Three patterns stand out:

• Freshness beats archive: the index is optimized for current content. Pages with a visible last-updated date and maintained facts are preferred.
• Many sources, but precise: Perplexity cites far more sources than other systems: an analysis of more than 118,000 AI answers found an average of around 22 citations per question, roughly three times ChatGPT. Pages that merely say "something about the topic" still get dropped.
• Diverse source types: alongside classic websites, Perplexity noticeably often cites specialist articles, documentation and community content.

Which content does Perplexity cite preferentially?

• Direct answers: the first paragraph answers the question — the rest goes deeper.
• Numbers with source and date: Perplexity backs claims; unsourced statements are hard to cite.
• Question headings: H2/H3 phrased as questions make the question-to-passage mapping trivial.
• Structured blocks: lists, tables and FAQ sections extract cleanly.
• Maintained freshness: a visible "as of …" signals that the facts are checked.

What role do PerplexityBot and Perplexity-User play?

Perplexity uses two kinds of access (Perplexity crawler documentation):

• PerplexityBot crawls for the search index and respects robots.txt. Blocked pages are indexed without full text; per Perplexity’s help center the domain, headline and a brief factual summary can still appear — but that does not make a substantive citation. According to Perplexity, the bot is not used to train foundation models.
• Perplexity-User fetches pages live when users ask a concrete question. Perplexity treats this access as an agent acting on behalf of users that does not strictly follow robots.txt.

In practice this means: allow PerplexityBot in robots.txt, serve content server-side (no JavaScript-only rendering) and offer curated context via llms.txt.

How do you check whether you are cited?

• Ask yourself: put your audience's typical questions to Perplexity and check the source list — including phrasing variants.
• Referral traffic: watch for visits with perplexity.ai as the referrer in your analytics.
• Systematic monitoring: check the same questions automatically on a schedule and track mentions over time — that is how the visibility check in flize does it.

Perplexity vs. ChatGPT: the differences

The criteria overlap heavily — if you optimize for ChatGPT, you are well positioned for Perplexity too. Three differences are worth knowing:

• Citation by default: ChatGPT cites only in search-backed answers, Perplexity always. Every question awards source slots.
• Freshness: Perplexity's index is closer to the news cycle; maintained, dated content gets picked up faster.
• Own index: Perplexity crawls itself and runs its own search index rather than relying on Google or Bing, supplemented by third-party crawler data. That is exactly what makes PerplexityBot access all the more important.

It becomes systematic when you continuously know what your audience is asking right now, check every question against your website and close the gaps with answer-ready articles. That is exactly the loop flize automates: scouting, answer check, content production and attribution.