How to rank in Perplexity: crawlers, server HTML and citable passages
Let PerplexityBot through robots.txt and Cloudflare, keep facts out of JS-only components and earn spots on recurring sources.
·3 min read
Learning how to rank in Perplexity comes down to three practical steps: allow Perplexity's crawlers through robots.txt and your CDN, keep key answers in server-rendered HTML as standalone passages, and earn mentions on the third-party roundups Perplexity already cites. Because Perplexity scores individual sub-document passages rather than whole pages, a concise answer in plain HTML frequently beats a long guide that hides facts behind scripts.
How Perplexity picks pages to cite
Perplexity queries an index fed by PerplexityBot across hundreds of billions of webpages, breaks pages into passage-sized units and scores each passage against the question. A single buyer prompt in Pro Search often triggers several smaller searches (pricing, integrations, reviews), giving different pages a chance to earn numbered citations.
Some cited domains carry Government, Academic or Trusted badges; Perplexity states payments and partnerships do not buy these labels and most sites carry no label at all.
Allow PerplexityBot without breaking disallows
PerplexityBot crawls for the search index and obeys robots.txt, while Perplexity-User fetches live pages for a user's question and generally ignores robots.txt. Neither trains foundation models according to Perplexity's crawler docs.
When adding a PerplexityBot block in robots.txt, repeat every Disallow rule from User-agent: *, because a bot reads only the most specific matching group:
User-agent: PerplexityBot
Allow: /
Disallow: /cart/
Disallow: /account/
User-agent: *
Disallow: /cart/
Disallow: /account/
Changes can take up to 24 hours to apply. Check Cloudflare or your WAF too: on September 15, 2026 Cloudflare replaced its old AI toggle with separate Search, Training and Agent controls (plus AI Crawl Control), and ad-supported sites often start with Agent blocked. Verify in server logs that PerplexityBot requests from Perplexity's published IP JSON ranges return HTTP 200, not 403.
Keep prices and specs in server HTML
Major AI crawlers, including Perplexity's, do not execute JavaScript when fetching pages, as Vercel's crawler study showed. Any pricing table, spec sheet or FAQ injected client-side is invisible to PerplexityBot. Test with curl against a pricing page or compare raw HTML against rendered HTML in seogeoaeo.ai's Perplexity SEO guide and AI Crawler View.
Write passages that stand alone
| Weak passage | Citable passage |
|---|---|
| Uses vague pronouns like "it" or "our tool" | Names the product and plan explicitly |
| Buries numbers under intro paragraphs | Puts the question in a heading and answers in sentence one |
| Hides updates or fakes daily dates | Displays an accurate publish and last-updated date |
Because Perplexity extracts single passages, every section should make sense if read on its own.
Earn mentions on Perplexity's recurring sources
Ask 10 to 20 real buyer questions in Perplexity and note which third-party domains keep showing up: comparison sites, forums, reviews and videos. To check multiple engines systematically, seogeoaeo.ai offers AI Answer Visibility (120 credits per run as of November 2026), which queries ChatGPT, Perplexity and Gemini with web search enabled and shows your brand's rank, cited URLs and which competitor sites appear instead.
Measure referrals and live fetches
Check perplexity.ai under GA4 Traffic acquisition (Session source), keeping in mind app clicks often arrive with no referrer as Direct traffic. Pair that floor with server logs counting Perplexity-User hits by URL to spot which pages live questions pull in.
FAQ
Does blocking PerplexityBot remove my site from Perplexity completely?
No. Perplexity's documentation states it skips indexing page text for blocked URLs, yet may still list the domain, headline and a brief summary, which leaves Perplexity with no passages of yours to cite.
Why do Cloudflare sites block Perplexity even with robots.txt set to Allow?
Cloudflare's Search, Training and Agent security controls (introduced September 15, 2026) and AI Crawl Control can return HTTP 403 upstream of your origin server, especially on ad-supported domains where Agent defaults to Block.
Does Perplexity train foundation models on crawled pages?
Perplexity's crawler documentation states that neither PerplexityBot nor Perplexity-User trains AI foundation models.