Perplexity SEO: how Perplexity finds, reads and cites pages
Perplexity searches the web for each question and links the pages it used as numbered citations. To be one of them, let PerplexityBot crawl your site, since it feeds Perplexity's search results, and make sure your firewall lets Perplexity's crawlers through. Perplexity publishes no ranking factors. In our runs, most of the sites it cited for a question were sites no other engine cited for it.
Checked against the studies and help pages it cites on October 20, 2026.
How Perplexity finds and reads pages.
Perplexity's help pages describe the path from a question to a cited page in outline. They leave out how it ranks what it finds.
A search for every question
Perplexity searches the web in real time for a question and sums up what it finds, and each answer carries numbered citations that link to the original sources.[1]
Two crawlers, two jobs
PerplexityBot surfaces and links websites in Perplexity's search results. Perplexity-User visits a page when a user's question calls for it. Neither, by Perplexity's account, collects content to train AI models.[2] Their pages have the user agents, IP lists and robots.txt names.
Pro Search reads more
Pro Search runs several searches and draws on dozens of sources, where Perplexity's own table gives a standard search one or two. Deep Research runs dozens of searches and reads hundreds of sources for a report.[1, 3]
The writer changes
In Pro Search, its default picks the model for each question, and Pro users can choose Perplexity's own Sonar or models such as GPT-5, Claude Sonnet 4.6 and Gemini 3.1 Pro. So the same pages can be summed up differently from one person to the next.[1, 3]
No published ranking factors
The help page says Perplexity gathers from authoritative sources such as articles, websites and journals, and stops there.[1] We found no Perplexity page that lists ranking factors or a way to submit a site.
Letting Perplexity in, and the dispute over it.
Getting in is the part Perplexity documents. How its fetcher treats sites that say no is disputed.
Allow PerplexityBot, and its IP addresses
For a site to appear in its search results, Perplexity says to allow PerplexityBot in robots.txt and permit requests from its published IP ranges. Behind a web application firewall, a rule may need to allow its bots by user agent and IP address together. A robots.txt change can take up to 24 hours to register.[2] The firewall guide shows where each host keeps that setting.
Perplexity-User and robots.txt
Perplexity says that since a user asked for the fetch, Perplexity-User generally ignores robots.txt rules.[2]
Cloudflare's test
In August 2025 Cloudflare reported that when it blocked Perplexity's declared crawlers on new domains whose robots.txt shut out every bot, Perplexity still answered with details of their content, and requests came from a browser identity posing as Chrome on a Mac, from addresses outside Perplexity's lists. It counted 3 to 6 million such requests a day, against 20 to 25 million from the declared fetcher, and removed Perplexity from its verified bots.[4]
Perplexity's reply
Perplexity answered the same day that Cloudflare had confused its traffic with BrowserBase, a third-party cloud browser service it uses occasionally for specialized tasks, and put its own use of it under 45,000 requests a day. It argues that an agent fetching a page for a person's request is not a bot.[5]
If you want to stay out
A Disallow line alone may not do it for user-requested fetches. Only a firewall rule turns away a fetcher that doesn't follow robots.txt, and the firewall guide shows how.
Perplexity beside five other engines.
We compared what Perplexity cited with what five other engines cited for the same question in the same run, and checked whether it linked a brand it named.
of the sites Perplexity cited for a question (1,459 of 2,019) were cited by none of the other five engines for that question in that run.
questions where Perplexity and ChatGPT cited at least one site in common.
Perplexity answers that named Plausible also linked a page on plausible.io, of 66 answers to questions that don't name it.
answers cited Reddit, and 0 cited YouTube, through Perplexity's API. Studies of what Perplexity cites find both near the top.
The answers are from GEOFast's two demo projects (4 runs for trygeo.fast and 3 for plausible.io on October 6 and 7, 2026), on questions about AI visibility tools and about web analytics, in US English: 191 Perplexity answers. Perplexity was asked through its API with its search, which always searches; the other engines were ChatGPT, Claude, Gemini and Google's AI Overviews and AI Mode. A site counts once per answer. The Plausible figure leaves out questions that name the brand.
Limits: the API's answers can differ from what the Perplexity app shows. Two fields and two days; GEOFast's onboarding drafted the questions, so they aren't measured demand.
What studies show about Perplexity.
Few studies look at Perplexity alone. These are the ones with a method you can check.
Quotes and statistics, tested on Perplexity
The GEO paper tested its methods on Perplexity itself. Adding quotations did best, a 22% gain on its position-adjusted word count; adding statistics and citing sources gained up to 9% and 37% on its two measures; keyword stuffing did 10% worse than leaving the page alone.[6]
Reddit and YouTube lead its citations
Across 3.1 million US queries in September 2026, Reddit took 21.6% of the citations that went to Perplexity's 50 most-cited sites, YouTube 20.8% and Wikipedia 6.3%. The data comes from Brand Radar, which Ahrefs sells.[7] Our API runs, on far fewer questions, saw little of either.
No JavaScript
Vercel's 2024 study of AI crawlers' traffic listed PerplexityBot among the crawlers that don't run JavaScript.[8] The JavaScript guide covers what that means for your pages.
Perplexity SEO, step by step.
Access is the part you control fully. The rest raises your odds; nothing in Perplexity's pages guarantees a citation.
Let both crawlers through
Allow PerplexityBot in robots.txt and in your firewall, and let Perplexity-User's addresses through so answers can read a page on request. The free AI crawler checker reads your robots.txt for both and fetches your homepage as PerplexityBot, which shows a firewall block that robots.txt can't.
Put the text in the HTML
The JavaScript rendering checker shows how much of a page's text only appears once scripts run.
Give it facts to quote
Quotations and statistics are what helped most in the GEO paper's tests on Perplexity. A page that states its numbers plainly, with their source, gives an answer something to lift.
Look at what it cites for your questions
Since most of the sites Perplexity cited in our runs were cited by no other engine for the same question, its sources are worth reading on their own. Ask your buyers' questions, open the citations, and see which lists, reviews and guides it relies on.
ChatGPT is a separate job
ChatGPT has its own crawler and its own sources: see getting cited by ChatGPT.
Counted for you
AI visibility tracking asks Perplexity your questions every week, with the sites it cited, and shows each rate with its 90% range.
Questions about Perplexity SEO.
Can you submit a site to Perplexity?
We found no Perplexity page that describes a way to submit a site or a page. What its crawler documentation asks for is access: allow PerplexityBot in robots.txt and let requests from its published IP addresses through, so your site can appear in its search results.
Is Perplexity SEO different from SEO for ChatGPT?
The groundwork is the same: crawlers that can reach the page, and text in the HTML. The crawler and the sources differ. Perplexity has its own, PerplexityBot, and in our runs 72% of the sites it cited for a question were cited by no other engine we asked, ChatGPT included.
Does Perplexity cite Reddit a lot?
In Ahrefs' September 2026 data from 3.1 million US queries, Reddit took 21.6% of the citations that went to Perplexity's 50 most-cited sites, with YouTube at 20.8%. In our runs through Perplexity's API, Reddit was cited in 7 of 191 answers, so what the API cites can differ from what Ahrefs saw.
Did Perplexity crawl sites that blocked it?
Cloudflare said so in August 2025: when it blocked Perplexity's declared crawlers on test sites closed by robots.txt, it saw requests from an undeclared browser identity and could still get details of the sites from Perplexity. Perplexity replied that Cloudflare had confused its traffic with a third-party cloud browser service it uses occasionally.
Which AI model writes Perplexity's answers?
It depends on the person. In Pro Search, Perplexity's default picks a model for each question, and Pro users can choose its own Sonar or models such as GPT-5, Claude Sonnet 4.6 and Gemini 3.1 Pro. The pages it searched are written up by whichever model is set.
Where each fact comes from.
Each numbered source is the page a fact comes from: published studies and the AI companies' own help pages, read on October 20, 2026. The answers above repeat facts marked in the guide. Help pages change, so check them before you act.
- 1Perplexity: How does Perplexity work?perplexity.ai/help-center/en/articles/10352895-how-does-perplexity-work
- 2Perplexity: Perplexity Crawlersdocs.perplexity.ai/docs/resources/perplexity-crawlers
- 3Perplexity: What is Pro Search?perplexity.ai/help-center/en/articles/10352903-what-is-pro-search
- 4Cloudflare: Perplexity is using stealth, undeclared crawlers to evade website no-crawl directivesblog.cloudflare.com/perplexity-is-using-stealth-undeclared-crawlers-to-evade-website-no-crawl-directives/
- 5Perplexity: Agents or bots? Making sense of AI on the open webperplexity.ai/hub/blog/agents-or-bots-making-sense-of-ai-on-the-open-web
- 6Aggarwal et al., KDD 2024: GEO: Generative Engine Optimizationarxiv.org/abs/2311.09735
- 7Ahrefs: The 50 Most-Cited Websites in Perplexity (September 2026)ahrefs.com/blog/most-cited-domains-perplexity/
- 8Vercel: The rise of the AI crawlervercel.com/blog/the-rise-of-the-ai-crawler
Keep reading
- The free AI crawler checkerTest 13 AI crawlers, from GPTBot to PerplexityBot, against your robots.txt, and fetch your homepage as 4 of them. Free, no sign-up.
- AI visibility tracking across six AI enginesTrack how often ChatGPT, AI Overviews, AI Mode, Gemini, Perplexity and Claude name you, put you first and cite your site, with a 90% range on every rate.
- Perplexity rank tracking, weekly, with rangesTrack how often Perplexity names you, puts you first and cites your site, asked weekly through its Sonar API with search, with a 90% range on each rate.
- AI visibility tools, comparedEvery tool we have checked, by starting price.
- What is GEOGEO, generative engine optimization, in plain words: how AI answers pick their sources, what research shows helps, and how to measure your AI visibility.
- AEO vs GEO vs SEOAEO vs GEO vs SEO: where each name came from, what Google and Microsoft say, the work all three share, and the four things that change with AI answers.
- Check if ChatGPT mentions your brandCheck by hand in five steps, see why one chat can't measure it (answers change, memory, location, leading questions), and where ChatGPT and Claude do help.
- Find the competitors ChatGPT recommendsA method for finding who ChatGPT recommends instead of you: which questions to ask, how many times, how to count the names and work out your share of voice.
- Does ChatGPT give everyone the same answer?No. ChatGPT writes each answer afresh and can use your memory, location, a web search and your plan's model. OpenAI's own pages, and our repeat-run numbers.
- AI visibility metricsMention rate, first-pick rate, citation rate, share of voice and sentiment: what each one counts, how to work it out, and why each needs a range.
- Get ChatGPT to recommend your businessWhat gets a business named in ChatGPT's answers: what you control, whether you can submit your business, how long changes take, and how to measure it honestly.
- Get cited by ChatGPTHow ChatGPT search finds and picks the pages it links, from OpenAI's docs, what makes a page usable as a source, and which kinds of pages our data shows cited.
- Why ChatGPT doesn't show your websiteSix reasons ChatGPT leaves a site out, in the order to check them: robots.txt, firewalls, noindex, JavaScript, Bing's index and who the pages it reads name.
- Does ChatGPT use Google or Bing?What OpenAI documents about where ChatGPT search gets results (Bing, its own crawler and index), what is only reported about Google, and what it means for you.
- Rank in Google's AI OverviewsWhat Google says you need to show up in AI Overviews, how query fan-out and nosnippet work, and what studies show: 38% of cited pages rank in the top 10.
- Rank in Google AI ModeWhat sets Google's AI Mode apart from AI Overviews: query fan-out, follow-ups, a custom Gemini and Search Console's counts, and how often the two cite one page.
- Which AI crawlers to allowWhich AI crawlers feed ChatGPT, Claude, Perplexity and Google's answers, which only train models, and what the 1,529 of 1,605 cited sites we could read allow.
- robots.txt for AI crawlersCopy-ready robots.txt rules that keep AI search crawlers in and training crawlers out, how each crawler reads the file, and the mistakes that lock AI out.
- What is llms.txt?llms.txt is a Markdown file that tells AI agents what a site holds. Its format, a real example, who reads it today, and what studies show about its effect.
- Firewalls that block AI crawlersCloudflare, Vercel, AWS WAF and others can turn AI crawlers away even when robots.txt lets them in. Where each setting lives and how to let crawlers through.
- Do AI crawlers render JavaScript?Google's and Bing's crawlers render JavaScript. OpenAI, Anthropic and Perplexity don't say, and Vercel's 2024 study found theirs don't. Our data on cited pages.
Find out what AI says about your site.
Paste your URL and leave with a ranked list of what to fix, with steps and code.
- Free
- No account